@namzu/sdk 2.0.0 → 3.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +2265 -0
- package/README.md +49 -53
- package/dist/__tests__/default-path-conformance.test.d.ts +2 -0
- package/dist/__tests__/default-path-conformance.test.d.ts.map +1 -0
- package/dist/__tests__/default-path-conformance.test.js +262 -0
- package/dist/__tests__/default-path-conformance.test.js.map +1 -0
- package/dist/__tests__/shipped-defaults.test.d.ts +2 -0
- package/dist/__tests__/shipped-defaults.test.d.ts.map +1 -0
- package/dist/__tests__/shipped-defaults.test.js +97 -0
- package/dist/__tests__/shipped-defaults.test.js.map +1 -0
- package/dist/advisory/__tests__/budget-enforcement.test.d.ts +2 -0
- package/dist/advisory/__tests__/budget-enforcement.test.d.ts.map +1 -0
- package/dist/advisory/__tests__/budget-enforcement.test.js +159 -0
- package/dist/advisory/__tests__/budget-enforcement.test.js.map +1 -0
- package/dist/advisory/__tests__/structured-response.test.d.ts +2 -0
- package/dist/advisory/__tests__/structured-response.test.d.ts.map +1 -0
- package/dist/advisory/__tests__/structured-response.test.js +99 -0
- package/dist/advisory/__tests__/structured-response.test.js.map +1 -0
- package/dist/advisory/budget.d.ts +15 -0
- package/dist/advisory/budget.d.ts.map +1 -0
- package/dist/advisory/budget.js +21 -0
- package/dist/advisory/budget.js.map +1 -0
- package/dist/advisory/context.d.ts +2 -0
- package/dist/advisory/context.d.ts.map +1 -1
- package/dist/advisory/context.js +16 -0
- package/dist/advisory/context.js.map +1 -1
- package/dist/advisory/executor.d.ts +14 -11
- package/dist/advisory/executor.d.ts.map +1 -1
- package/dist/advisory/executor.js +36 -24
- package/dist/advisory/executor.js.map +1 -1
- package/dist/advisory/executor.test.d.ts +8 -6
- package/dist/advisory/executor.test.d.ts.map +1 -1
- package/dist/advisory/executor.test.js +12 -8
- package/dist/advisory/executor.test.js.map +1 -1
- package/dist/advisory/index.d.ts +2 -0
- package/dist/advisory/index.d.ts.map +1 -1
- package/dist/advisory/index.js +2 -0
- package/dist/advisory/index.js.map +1 -1
- package/dist/advisory/parse.d.ts +23 -0
- package/dist/advisory/parse.d.ts.map +1 -0
- package/dist/advisory/parse.js +56 -0
- package/dist/advisory/parse.js.map +1 -0
- package/dist/agents/AbstractAgent.d.ts +48 -0
- package/dist/agents/AbstractAgent.d.ts.map +1 -1
- package/dist/agents/AbstractAgent.js +77 -0
- package/dist/agents/AbstractAgent.js.map +1 -1
- package/dist/agents/PipelineAgent.d.ts +10 -0
- package/dist/agents/PipelineAgent.d.ts.map +1 -1
- package/dist/agents/PipelineAgent.js +12 -0
- package/dist/agents/PipelineAgent.js.map +1 -1
- package/dist/agents/ReactiveAgent.d.ts +10 -0
- package/dist/agents/ReactiveAgent.d.ts.map +1 -1
- package/dist/agents/ReactiveAgent.js +40 -0
- package/dist/agents/ReactiveAgent.js.map +1 -1
- package/dist/agents/RouterAgent.d.ts +10 -0
- package/dist/agents/RouterAgent.d.ts.map +1 -1
- package/dist/agents/RouterAgent.js +23 -2
- package/dist/agents/RouterAgent.js.map +1 -1
- package/dist/agents/SupervisorAgent.d.ts +11 -1
- package/dist/agents/SupervisorAgent.d.ts.map +1 -1
- package/dist/agents/SupervisorAgent.js +23 -1
- package/dist/agents/SupervisorAgent.js.map +1 -1
- package/dist/agents/__tests__/idempotency-key.test.d.ts +2 -0
- package/dist/agents/__tests__/idempotency-key.test.d.ts.map +1 -0
- package/dist/agents/__tests__/idempotency-key.test.js +149 -0
- package/dist/agents/__tests__/idempotency-key.test.js.map +1 -0
- package/dist/agents/__tests__/invocation-lock.test.d.ts +2 -0
- package/dist/agents/__tests__/invocation-lock.test.d.ts.map +1 -0
- package/dist/agents/__tests__/invocation-lock.test.js +126 -0
- package/dist/agents/__tests__/invocation-lock.test.js.map +1 -0
- package/dist/agents/__tests__/reachability.test.d.ts +2 -0
- package/dist/agents/__tests__/reachability.test.d.ts.map +1 -0
- package/dist/agents/__tests__/reachability.test.js +173 -0
- package/dist/agents/__tests__/reachability.test.js.map +1 -0
- package/dist/agents/__tests__/supervisor-ledger.test.js +1 -1
- package/dist/agents/__tests__/supervisor-ledger.test.js.map +1 -1
- package/dist/agents/index.d.ts +1 -1
- package/dist/agents/index.d.ts.map +1 -1
- package/dist/agents/lock.d.ts +0 -1
- package/dist/agents/lock.d.ts.map +1 -1
- package/dist/agents/lock.js.map +1 -1
- package/dist/bridge/a2a/late-events.test.d.ts +2 -0
- package/dist/bridge/a2a/late-events.test.d.ts.map +1 -0
- package/dist/bridge/a2a/late-events.test.js +70 -0
- package/dist/bridge/a2a/late-events.test.js.map +1 -0
- package/dist/bridge/a2a/mapper.d.ts.map +1 -1
- package/dist/bridge/a2a/mapper.js +55 -3
- package/dist/bridge/a2a/mapper.js.map +1 -1
- package/dist/bridge/a2a/mapper.test.js +1 -1
- package/dist/bridge/a2a/mapper.test.js.map +1 -1
- package/dist/bridge/a2a/message.d.ts.map +1 -1
- package/dist/bridge/a2a/message.js +5 -1
- package/dist/bridge/a2a/message.js.map +1 -1
- package/dist/bridge/sse/late-events.test.d.ts +2 -0
- package/dist/bridge/sse/late-events.test.d.ts.map +1 -0
- package/dist/bridge/sse/late-events.test.js +163 -0
- package/dist/bridge/sse/late-events.test.js.map +1 -0
- package/dist/bridge/sse/mapper.d.ts.map +1 -1
- package/dist/bridge/sse/mapper.js +100 -0
- package/dist/bridge/sse/mapper.js.map +1 -1
- package/dist/compaction/__tests__/cleared-result-recovery.test.d.ts +2 -0
- package/dist/compaction/__tests__/cleared-result-recovery.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/cleared-result-recovery.test.js +101 -0
- package/dist/compaction/__tests__/cleared-result-recovery.test.js.map +1 -0
- package/dist/compaction/__tests__/context-window.test.d.ts +2 -0
- package/dist/compaction/__tests__/context-window.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/context-window.test.js +62 -0
- package/dist/compaction/__tests__/context-window.test.js.map +1 -0
- package/dist/compaction/__tests__/dangling.test.js +16 -5
- package/dist/compaction/__tests__/dangling.test.js.map +1 -1
- package/dist/compaction/__tests__/lossiness.test.d.ts +2 -0
- package/dist/compaction/__tests__/lossiness.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/lossiness.test.js +163 -0
- package/dist/compaction/__tests__/lossiness.test.js.map +1 -0
- package/dist/compaction/__tests__/state-across-resume.test.d.ts +2 -0
- package/dist/compaction/__tests__/state-across-resume.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/state-across-resume.test.js +88 -0
- package/dist/compaction/__tests__/state-across-resume.test.js.map +1 -0
- package/dist/compaction/__tests__/tool-result-editing.test.d.ts +2 -0
- package/dist/compaction/__tests__/tool-result-editing.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/tool-result-editing.test.js +184 -0
- package/dist/compaction/__tests__/tool-result-editing.test.js.map +1 -0
- package/dist/compaction/__tests__/trim-opens-on-user.test.d.ts +2 -0
- package/dist/compaction/__tests__/trim-opens-on-user.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/trim-opens-on-user.test.js +109 -0
- package/dist/compaction/__tests__/trim-opens-on-user.test.js.map +1 -0
- package/dist/compaction/__tests__/verifier-model.test.d.ts +2 -0
- package/dist/compaction/__tests__/verifier-model.test.d.ts.map +1 -0
- package/dist/compaction/__tests__/verifier-model.test.js +58 -0
- package/dist/compaction/__tests__/verifier-model.test.js.map +1 -0
- package/dist/compaction/context-window.d.ts +44 -0
- package/dist/compaction/context-window.d.ts.map +1 -0
- package/dist/compaction/context-window.js +97 -0
- package/dist/compaction/context-window.js.map +1 -0
- package/dist/compaction/dangling.js +48 -1
- package/dist/compaction/dangling.js.map +1 -1
- package/dist/compaction/extractor.d.ts.map +1 -1
- package/dist/compaction/extractor.js +55 -1
- package/dist/compaction/extractor.js.map +1 -1
- package/dist/compaction/index.d.ts +5 -0
- package/dist/compaction/index.d.ts.map +1 -1
- package/dist/compaction/index.js +3 -0
- package/dist/compaction/index.js.map +1 -1
- package/dist/compaction/manager.d.ts +25 -0
- package/dist/compaction/manager.d.ts.map +1 -1
- package/dist/compaction/manager.js +42 -8
- package/dist/compaction/manager.js.map +1 -1
- package/dist/compaction/managers/structured.d.ts.map +1 -1
- package/dist/compaction/managers/structured.js +7 -3
- package/dist/compaction/managers/structured.js.map +1 -1
- package/dist/compaction/retention.d.ts +15 -0
- package/dist/compaction/retention.d.ts.map +1 -0
- package/dist/compaction/retention.js +89 -0
- package/dist/compaction/retention.js.map +1 -0
- package/dist/compaction/serializer.d.ts.map +1 -1
- package/dist/compaction/serializer.js +28 -11
- package/dist/compaction/serializer.js.map +1 -1
- package/dist/compaction/tool-result-editing.d.ts +54 -0
- package/dist/compaction/tool-result-editing.d.ts.map +1 -0
- package/dist/compaction/tool-result-editing.js +164 -0
- package/dist/compaction/tool-result-editing.js.map +1 -0
- package/dist/compaction/types.d.ts +10 -0
- package/dist/compaction/types.d.ts.map +1 -1
- package/dist/compaction/verifier.d.ts +20 -1
- package/dist/compaction/verifier.d.ts.map +1 -1
- package/dist/compaction/verifier.js +11 -2
- package/dist/compaction/verifier.js.map +1 -1
- package/dist/compaction/wire.d.ts +54 -0
- package/dist/compaction/wire.d.ts.map +1 -0
- package/dist/compaction/wire.js +45 -0
- package/dist/compaction/wire.js.map +1 -0
- package/dist/config/runtime.d.ts +161 -45
- package/dist/config/runtime.d.ts.map +1 -1
- package/dist/config/runtime.js +61 -19
- package/dist/config/runtime.js.map +1 -1
- package/dist/connector/builtins/http.d.ts.map +1 -1
- package/dist/connector/builtins/http.js +14 -1
- package/dist/connector/builtins/http.js.map +1 -1
- package/dist/connector/builtins/http.test.js +12 -0
- package/dist/connector/builtins/http.test.js.map +1 -1
- package/dist/connector/mcp/__tests__/discovery-policy.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/discovery-policy.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/discovery-policy.test.js +109 -0
- package/dist/connector/mcp/__tests__/discovery-policy.test.js.map +1 -0
- package/dist/connector/mcp/__tests__/list-pagination.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/list-pagination.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/list-pagination.test.js +81 -0
- package/dist/connector/mcp/__tests__/list-pagination.test.js.map +1 -0
- package/dist/connector/mcp/__tests__/policy.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/policy.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/policy.test.js +86 -0
- package/dist/connector/mcp/__tests__/policy.test.js.map +1 -0
- package/dist/connector/mcp/__tests__/protocol-negotiation.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/protocol-negotiation.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/protocol-negotiation.test.js +81 -0
- package/dist/connector/mcp/__tests__/protocol-negotiation.test.js.map +1 -0
- package/dist/connector/mcp/__tests__/schema-fidelity.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/schema-fidelity.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/schema-fidelity.test.js +206 -0
- package/dist/connector/mcp/__tests__/schema-fidelity.test.js.map +1 -0
- package/dist/connector/mcp/__tests__/structured-result.test.d.ts +2 -0
- package/dist/connector/mcp/__tests__/structured-result.test.d.ts.map +1 -0
- package/dist/connector/mcp/__tests__/structured-result.test.js +112 -0
- package/dist/connector/mcp/__tests__/structured-result.test.js.map +1 -0
- package/dist/connector/mcp/adapter.d.ts +14 -0
- package/dist/connector/mcp/adapter.d.ts.map +1 -1
- package/dist/connector/mcp/adapter.js +292 -18
- package/dist/connector/mcp/adapter.js.map +1 -1
- package/dist/connector/mcp/client.d.ts +38 -0
- package/dist/connector/mcp/client.d.ts.map +1 -1
- package/dist/connector/mcp/client.js +138 -12
- package/dist/connector/mcp/client.js.map +1 -1
- package/dist/connector/mcp/client.test.d.ts +2 -0
- package/dist/connector/mcp/client.test.d.ts.map +1 -0
- package/dist/connector/mcp/client.test.js +105 -0
- package/dist/connector/mcp/client.test.js.map +1 -0
- package/dist/connector/mcp/discovery.d.ts +32 -1
- package/dist/connector/mcp/discovery.d.ts.map +1 -1
- package/dist/connector/mcp/discovery.js +41 -4
- package/dist/connector/mcp/discovery.js.map +1 -1
- package/dist/connector/mcp/http-sse.d.ts.map +1 -1
- package/dist/connector/mcp/http-sse.js +15 -1
- package/dist/connector/mcp/http-sse.js.map +1 -1
- package/dist/connector/mcp/index.d.ts +3 -0
- package/dist/connector/mcp/index.d.ts.map +1 -1
- package/dist/connector/mcp/index.js +1 -0
- package/dist/connector/mcp/index.js.map +1 -1
- package/dist/connector/mcp/policy.d.ts +57 -0
- package/dist/connector/mcp/policy.d.ts.map +1 -0
- package/dist/connector/mcp/policy.js +84 -0
- package/dist/connector/mcp/policy.js.map +1 -0
- package/dist/connector/mcp/schema-refs.d.ts +26 -0
- package/dist/connector/mcp/schema-refs.d.ts.map +1 -0
- package/dist/connector/mcp/schema-refs.js +111 -0
- package/dist/connector/mcp/schema-refs.js.map +1 -0
- package/dist/constants/agent/index.d.ts.map +1 -1
- package/dist/constants/agent/index.js +3 -0
- package/dist/constants/agent/index.js.map +1 -1
- package/dist/constants/compaction/index.js +1 -1
- package/dist/constants/compaction/index.js.map +1 -1
- package/dist/constants/continuation.d.ts +4 -4
- package/dist/constants/continuation.js +4 -4
- package/dist/constants/mcp/index.d.ts +31 -0
- package/dist/constants/mcp/index.d.ts.map +1 -1
- package/dist/constants/mcp/index.js +31 -0
- package/dist/constants/mcp/index.js.map +1 -1
- package/dist/constants/plugin/index.d.ts +23 -2
- package/dist/constants/plugin/index.d.ts.map +1 -1
- package/dist/constants/plugin/index.js +23 -2
- package/dist/constants/plugin/index.js.map +1 -1
- package/dist/constants/sandbox/index.d.ts +2 -3
- package/dist/constants/sandbox/index.d.ts.map +1 -1
- package/dist/constants/sandbox/index.js +6 -5
- package/dist/constants/sandbox/index.js.map +1 -1
- package/dist/constants/telemetry/index.d.ts +7 -0
- package/dist/constants/telemetry/index.d.ts.map +1 -1
- package/dist/constants/telemetry/index.js +7 -0
- package/dist/constants/telemetry/index.js.map +1 -1
- package/dist/constants/tools/index.d.ts +16 -0
- package/dist/constants/tools/index.d.ts.map +1 -1
- package/dist/constants/tools/index.js +16 -0
- package/dist/constants/tools/index.js.map +1 -1
- package/dist/contracts/a2a.d.ts +94 -94
- package/dist/contracts/api.d.ts +40 -2
- package/dist/contracts/api.d.ts.map +1 -1
- package/dist/contracts/run-status.d.ts +17 -0
- package/dist/contracts/run-status.d.ts.map +1 -0
- package/dist/contracts/run-status.js +41 -0
- package/dist/contracts/run-status.js.map +1 -0
- package/dist/contracts/schemas.d.ts +6 -6
- package/dist/eval/__tests__/case-deadline.test.d.ts +2 -0
- package/dist/eval/__tests__/case-deadline.test.d.ts.map +1 -0
- package/dist/eval/__tests__/case-deadline.test.js +128 -0
- package/dist/eval/__tests__/case-deadline.test.js.map +1 -0
- package/dist/eval/__tests__/end-to-end.test.d.ts +2 -0
- package/dist/eval/__tests__/end-to-end.test.d.ts.map +1 -0
- package/dist/eval/__tests__/end-to-end.test.js +223 -0
- package/dist/eval/__tests__/end-to-end.test.js.map +1 -0
- package/dist/eval/__tests__/harness-honesty.test.d.ts +2 -0
- package/dist/eval/__tests__/harness-honesty.test.d.ts.map +1 -0
- package/dist/eval/__tests__/harness-honesty.test.js +85 -0
- package/dist/eval/__tests__/harness-honesty.test.js.map +1 -0
- package/dist/eval/__tests__/harness.test.d.ts +2 -0
- package/dist/eval/__tests__/harness.test.d.ts.map +1 -0
- package/dist/eval/__tests__/harness.test.js +207 -0
- package/dist/eval/__tests__/harness.test.js.map +1 -0
- package/dist/eval/__tests__/judge.test.d.ts +2 -0
- package/dist/eval/__tests__/judge.test.d.ts.map +1 -0
- package/dist/eval/__tests__/judge.test.js +255 -0
- package/dist/eval/__tests__/judge.test.js.map +1 -0
- package/dist/eval/__tests__/scorer-gates.test.d.ts +2 -0
- package/dist/eval/__tests__/scorer-gates.test.d.ts.map +1 -0
- package/dist/eval/__tests__/scorer-gates.test.js +136 -0
- package/dist/eval/__tests__/scorer-gates.test.js.map +1 -0
- package/dist/eval/experiment.d.ts +54 -0
- package/dist/eval/experiment.d.ts.map +1 -0
- package/dist/eval/experiment.js +263 -0
- package/dist/eval/experiment.js.map +1 -0
- package/dist/eval/from-run.d.ts +30 -0
- package/dist/eval/from-run.d.ts.map +1 -0
- package/dist/eval/from-run.js +46 -0
- package/dist/eval/from-run.js.map +1 -0
- package/dist/eval/index.d.ts +8 -0
- package/dist/eval/index.d.ts.map +1 -0
- package/dist/eval/index.js +5 -0
- package/dist/eval/index.js.map +1 -0
- package/dist/eval/judge.d.ts +63 -0
- package/dist/eval/judge.d.ts.map +1 -0
- package/dist/eval/judge.js +136 -0
- package/dist/eval/judge.js.map +1 -0
- package/dist/eval/scorers.d.ts +40 -0
- package/dist/eval/scorers.d.ts.map +1 -0
- package/dist/eval/scorers.js +151 -0
- package/dist/eval/scorers.js.map +1 -0
- package/dist/eval/types.d.ts +150 -0
- package/dist/eval/types.d.ts.map +1 -0
- package/dist/eval/types.js +2 -0
- package/dist/eval/types.js.map +1 -0
- package/dist/execution/local.d.ts.map +1 -1
- package/dist/execution/local.js +6 -1
- package/dist/execution/local.js.map +1 -1
- package/dist/execution/local.test.d.ts +2 -0
- package/dist/execution/local.test.d.ts.map +1 -0
- package/dist/execution/local.test.js +37 -0
- package/dist/execution/local.test.js.map +1 -0
- package/dist/gateway/__tests__/list-survives-eviction.test.d.ts +2 -0
- package/dist/gateway/__tests__/list-survives-eviction.test.d.ts.map +1 -0
- package/dist/gateway/__tests__/list-survives-eviction.test.js +102 -0
- package/dist/gateway/__tests__/list-survives-eviction.test.js.map +1 -0
- package/dist/gateway/local.d.ts +42 -2
- package/dist/gateway/local.d.ts.map +1 -1
- package/dist/gateway/local.js +110 -3
- package/dist/gateway/local.js.map +1 -1
- package/dist/manager/agent/__tests__/dispose-cancels.test.d.ts +2 -0
- package/dist/manager/agent/__tests__/dispose-cancels.test.d.ts.map +1 -0
- package/dist/manager/agent/__tests__/dispose-cancels.test.js +88 -0
- package/dist/manager/agent/__tests__/dispose-cancels.test.js.map +1 -0
- package/dist/manager/agent/__tests__/lifecycle.test.js +230 -0
- package/dist/manager/agent/__tests__/lifecycle.test.js.map +1 -1
- package/dist/manager/agent/lifecycle.d.ts +35 -0
- package/dist/manager/agent/lifecycle.d.ts.map +1 -1
- package/dist/manager/agent/lifecycle.js +117 -7
- package/dist/manager/agent/lifecycle.js.map +1 -1
- package/dist/manager/run/persistence.d.ts +90 -1
- package/dist/manager/run/persistence.d.ts.map +1 -1
- package/dist/manager/run/persistence.js +120 -0
- package/dist/manager/run/persistence.js.map +1 -1
- package/dist/plugin/__tests__/enable-contributions.test.js +3 -3
- package/dist/plugin/__tests__/enable-contributions.test.js.map +1 -1
- package/dist/plugin/__tests__/hook-ordering.test.d.ts +2 -0
- package/dist/plugin/__tests__/hook-ordering.test.d.ts.map +1 -0
- package/dist/plugin/__tests__/hook-ordering.test.js +227 -0
- package/dist/plugin/__tests__/hook-ordering.test.js.map +1 -0
- package/dist/plugin/__tests__/install-time-refusal.test.d.ts +2 -0
- package/dist/plugin/__tests__/install-time-refusal.test.d.ts.map +1 -0
- package/dist/plugin/__tests__/install-time-refusal.test.js +49 -0
- package/dist/plugin/__tests__/install-time-refusal.test.js.map +1 -0
- package/dist/plugin/__tests__/lifecycle.test.js +95 -68
- package/dist/plugin/__tests__/lifecycle.test.js.map +1 -1
- package/dist/plugin/lifecycle.d.ts +15 -1
- package/dist/plugin/lifecycle.d.ts.map +1 -1
- package/dist/plugin/lifecycle.js +67 -20
- package/dist/plugin/lifecycle.js.map +1 -1
- package/dist/plugin/loader.d.ts +15 -0
- package/dist/plugin/loader.d.ts.map +1 -1
- package/dist/plugin/loader.js +33 -1
- package/dist/plugin/loader.js.map +1 -1
- package/dist/probe/registry.d.ts.map +1 -1
- package/dist/probe/registry.js +19 -1
- package/dist/probe/registry.js.map +1 -1
- package/dist/probe/registry.test.d.ts +9 -3
- package/dist/probe/registry.test.d.ts.map +1 -1
- package/dist/probe/registry.test.js +44 -6
- package/dist/probe/registry.test.js.map +1 -1
- package/dist/provider/__tests__/mock-script.test.d.ts +2 -0
- package/dist/provider/__tests__/mock-script.test.d.ts.map +1 -0
- package/dist/provider/__tests__/mock-script.test.js +151 -0
- package/dist/provider/__tests__/mock-script.test.js.map +1 -0
- package/dist/provider/__tests__/retry-visibility.test.d.ts +2 -0
- package/dist/provider/__tests__/retry-visibility.test.d.ts.map +1 -0
- package/dist/provider/__tests__/retry-visibility.test.js +137 -0
- package/dist/provider/__tests__/retry-visibility.test.js.map +1 -0
- package/dist/provider/__tests__/retry.test.d.ts +2 -0
- package/dist/provider/__tests__/retry.test.d.ts.map +1 -0
- package/dist/provider/__tests__/retry.test.js +259 -0
- package/dist/provider/__tests__/retry.test.js.map +1 -0
- package/dist/provider/capabilities.d.ts +1 -0
- package/dist/provider/capabilities.d.ts.map +1 -1
- package/dist/provider/capabilities.js +2 -0
- package/dist/provider/capabilities.js.map +1 -1
- package/dist/provider/collect.d.ts +1 -1
- package/dist/provider/collect.js +1 -1
- package/dist/provider/collect.test.d.ts +1 -1
- package/dist/provider/collect.test.js +1 -1
- package/dist/provider/index.d.ts +3 -0
- package/dist/provider/index.d.ts.map +1 -1
- package/dist/provider/index.js +2 -0
- package/dist/provider/index.js.map +1 -1
- package/dist/provider/mock-register.d.ts.map +1 -1
- package/dist/provider/mock-register.js +5 -2
- package/dist/provider/mock-register.js.map +1 -1
- package/dist/provider/mock.d.ts +35 -5
- package/dist/provider/mock.d.ts.map +1 -1
- package/dist/provider/mock.js +104 -32
- package/dist/provider/mock.js.map +1 -1
- package/dist/provider/retry.d.ts +42 -0
- package/dist/provider/retry.d.ts.map +1 -0
- package/dist/provider/retry.js +161 -0
- package/dist/provider/retry.js.map +1 -0
- package/dist/public-runtime.d.ts +26 -7
- package/dist/public-runtime.d.ts.map +1 -1
- package/dist/public-runtime.js +50 -7
- package/dist/public-runtime.js.map +1 -1
- package/dist/public-types.d.ts +5 -1
- package/dist/public-types.d.ts.map +1 -1
- package/dist/rag/__tests__/retrieval-math.test.d.ts +2 -0
- package/dist/rag/__tests__/retrieval-math.test.d.ts.map +1 -0
- package/dist/rag/__tests__/retrieval-math.test.js +46 -0
- package/dist/rag/__tests__/retrieval-math.test.js.map +1 -0
- package/dist/rag/chunking.d.ts +12 -0
- package/dist/rag/chunking.d.ts.map +1 -1
- package/dist/rag/chunking.js +32 -1
- package/dist/rag/chunking.js.map +1 -1
- package/dist/rag/embedding.d.ts +4 -4
- package/dist/rag/embedding.d.ts.map +1 -1
- package/dist/rag/embedding.js +23 -3
- package/dist/rag/embedding.js.map +1 -1
- package/dist/rag/embedding.test.d.ts +1 -1
- package/dist/rag/embedding.test.js +59 -10
- package/dist/rag/embedding.test.js.map +1 -1
- package/dist/rag/index.d.ts +1 -1
- package/dist/rag/index.d.ts.map +1 -1
- package/dist/rag/index.js +1 -1
- package/dist/rag/index.js.map +1 -1
- package/dist/rag/retriever.d.ts +0 -1
- package/dist/rag/retriever.d.ts.map +1 -1
- package/dist/rag/retriever.js +75 -22
- package/dist/rag/retriever.js.map +1 -1
- package/dist/rag/vector-store.test.d.ts +1 -1
- package/dist/rag/vector-store.test.js +2 -2
- package/dist/rag/vector-store.test.js.map +1 -1
- package/dist/registry/ManagedRegistry.test.d.ts +1 -1
- package/dist/registry/ManagedRegistry.test.js +1 -1
- package/dist/registry/Registry.test.d.ts +1 -1
- package/dist/registry/Registry.test.js +1 -1
- package/dist/registry/connector/definitions.test.d.ts +1 -1
- package/dist/registry/connector/definitions.test.js +1 -1
- package/dist/registry/tool/__tests__/schema.test.d.ts +2 -0
- package/dist/registry/tool/__tests__/schema.test.d.ts.map +1 -0
- package/dist/registry/tool/__tests__/schema.test.js +176 -0
- package/dist/registry/tool/__tests__/schema.test.js.map +1 -0
- package/dist/registry/tool/__tests__/tool-name-validation.test.d.ts +2 -0
- package/dist/registry/tool/__tests__/tool-name-validation.test.d.ts.map +1 -0
- package/dist/registry/tool/__tests__/tool-name-validation.test.js +70 -0
- package/dist/registry/tool/__tests__/tool-name-validation.test.js.map +1 -0
- package/dist/registry/tool/execute.d.ts +37 -0
- package/dist/registry/tool/execute.d.ts.map +1 -1
- package/dist/registry/tool/execute.js +91 -21
- package/dist/registry/tool/execute.js.map +1 -1
- package/dist/registry/tool/schema.d.ts +26 -0
- package/dist/registry/tool/schema.d.ts.map +1 -0
- package/dist/registry/tool/schema.js +77 -0
- package/dist/registry/tool/schema.js.map +1 -0
- package/dist/registry/toolset/catalog.d.ts.map +1 -1
- package/dist/registry/toolset/catalog.js +0 -1
- package/dist/registry/toolset/catalog.js.map +1 -1
- package/dist/registry/toolset/catalog.test.js +0 -1
- package/dist/registry/toolset/catalog.test.js.map +1 -1
- package/dist/run/reporter.d.ts.map +1 -1
- package/dist/run/reporter.js +76 -0
- package/dist/run/reporter.js.map +1 -1
- package/dist/runtime/bidi/__tests__/session.test.d.ts +2 -0
- package/dist/runtime/bidi/__tests__/session.test.d.ts.map +1 -0
- package/dist/runtime/bidi/__tests__/session.test.js +201 -0
- package/dist/runtime/bidi/__tests__/session.test.js.map +1 -0
- package/dist/runtime/bidi/index.d.ts +5 -0
- package/dist/runtime/bidi/index.d.ts.map +1 -0
- package/dist/runtime/bidi/index.js +3 -0
- package/dist/runtime/bidi/index.js.map +1 -0
- package/dist/runtime/bidi/mock.d.ts +34 -0
- package/dist/runtime/bidi/mock.d.ts.map +1 -0
- package/dist/runtime/bidi/mock.js +60 -0
- package/dist/runtime/bidi/mock.js.map +1 -0
- package/dist/runtime/bidi/session.d.ts +44 -0
- package/dist/runtime/bidi/session.d.ts.map +1 -0
- package/dist/runtime/bidi/session.js +154 -0
- package/dist/runtime/bidi/session.js.map +1 -0
- package/dist/runtime/decision/parser.d.ts.map +1 -1
- package/dist/runtime/decision/parser.js +8 -1
- package/dist/runtime/decision/parser.js.map +1 -1
- package/dist/runtime/decision/parser.test.d.ts +2 -0
- package/dist/runtime/decision/parser.test.d.ts.map +1 -0
- package/dist/runtime/decision/parser.test.js +37 -0
- package/dist/runtime/decision/parser.test.js.map +1 -0
- package/dist/runtime/query/__tests__/answer-review.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/answer-review.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/answer-review.test.js +141 -0
- package/dist/runtime/query/__tests__/answer-review.test.js.map +1 -0
- package/dist/runtime/query/__tests__/capability-negotiation.test.js +24 -111
- package/dist/runtime/query/__tests__/capability-negotiation.test.js.map +1 -1
- package/dist/runtime/query/__tests__/checkpoint-store.test.js +12 -44
- package/dist/runtime/query/__tests__/checkpoint-store.test.js.map +1 -1
- package/dist/runtime/query/__tests__/checkpoint.test.js +2 -10
- package/dist/runtime/query/__tests__/checkpoint.test.js.map +1 -1
- package/dist/runtime/query/__tests__/citations.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/citations.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/citations.test.js +85 -0
- package/dist/runtime/query/__tests__/citations.test.js.map +1 -0
- package/dist/runtime/query/__tests__/content-budget.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/content-budget.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/content-budget.test.js +69 -0
- package/dist/runtime/query/__tests__/content-budget.test.js.map +1 -0
- package/dist/runtime/query/__tests__/context.test.js +3 -2
- package/dist/runtime/query/__tests__/context.test.js.map +1 -1
- package/dist/runtime/query/__tests__/deferred-tools.test.js +21 -130
- package/dist/runtime/query/__tests__/deferred-tools.test.js.map +1 -1
- package/dist/runtime/query/__tests__/delegation-resume.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/delegation-resume.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/delegation-resume.test.js +111 -0
- package/dist/runtime/query/__tests__/delegation-resume.test.js.map +1 -0
- package/dist/runtime/query/__tests__/document-attachments.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/document-attachments.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/document-attachments.test.js +92 -0
- package/dist/runtime/query/__tests__/document-attachments.test.js.map +1 -0
- package/dist/runtime/query/__tests__/durable-park-and-trace.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/durable-park-and-trace.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/durable-park-and-trace.test.js +194 -0
- package/dist/runtime/query/__tests__/durable-park-and-trace.test.js.map +1 -0
- package/dist/runtime/query/__tests__/durable-question-park.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/durable-question-park.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/durable-question-park.test.js +253 -0
- package/dist/runtime/query/__tests__/durable-question-park.test.js.map +1 -0
- package/dist/runtime/query/__tests__/durable-run-state.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/durable-run-state.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/durable-run-state.test.js +158 -0
- package/dist/runtime/query/__tests__/durable-run-state.test.js.map +1 -0
- package/dist/runtime/query/__tests__/executor-plugin-hooks.test.js +24 -20
- package/dist/runtime/query/__tests__/executor-plugin-hooks.test.js.map +1 -1
- package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.js +187 -0
- package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.js.map +1 -0
- package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.js +95 -0
- package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.js.map +1 -0
- package/dist/runtime/query/__tests__/guardrails-e2e.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/guardrails-e2e.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/guardrails-e2e.test.js +132 -0
- package/dist/runtime/query/__tests__/guardrails-e2e.test.js.map +1 -0
- package/dist/runtime/query/__tests__/guardrails.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/guardrails.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/guardrails.test.js +163 -0
- package/dist/runtime/query/__tests__/guardrails.test.js.map +1 -0
- package/dist/runtime/query/__tests__/hook-model-visibility.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/hook-model-visibility.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/hook-model-visibility.test.js +119 -0
- package/dist/runtime/query/__tests__/hook-model-visibility.test.js.map +1 -0
- package/dist/runtime/query/__tests__/long-document-flow.test.js +9 -2
- package/dist/runtime/query/__tests__/long-document-flow.test.js.map +1 -1
- package/dist/runtime/query/__tests__/memory-promotion.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/memory-promotion.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/memory-promotion.test.js +119 -0
- package/dist/runtime/query/__tests__/memory-promotion.test.js.map +1 -0
- package/dist/runtime/query/__tests__/orphan-task-turn-end.test.js +5 -27
- package/dist/runtime/query/__tests__/orphan-task-turn-end.test.js.map +1 -1
- package/dist/runtime/query/__tests__/overflow-relief.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/overflow-relief.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/overflow-relief.test.js +119 -0
- package/dist/runtime/query/__tests__/overflow-relief.test.js.map +1 -0
- package/dist/runtime/query/__tests__/partial-batch-resume.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/partial-batch-resume.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/partial-batch-resume.test.js +205 -0
- package/dist/runtime/query/__tests__/partial-batch-resume.test.js.map +1 -0
- package/dist/runtime/query/__tests__/plugin-hooks.test.js +0 -1
- package/dist/runtime/query/__tests__/plugin-hooks.test.js.map +1 -1
- package/dist/runtime/query/__tests__/prepare-step-chain.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/prepare-step-chain.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/prepare-step-chain.test.js +104 -0
- package/dist/runtime/query/__tests__/prepare-step-chain.test.js.map +1 -0
- package/dist/runtime/query/__tests__/prepare-step.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/prepare-step.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/prepare-step.test.js +194 -0
- package/dist/runtime/query/__tests__/prepare-step.test.js.map +1 -0
- package/dist/runtime/query/__tests__/probe-veto-is-error.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/probe-veto-is-error.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/probe-veto-is-error.test.js +111 -0
- package/dist/runtime/query/__tests__/probe-veto-is-error.test.js.map +1 -0
- package/dist/runtime/query/__tests__/prompt.test.js +1 -1
- package/dist/runtime/query/__tests__/prompt.test.js.map +1 -1
- package/dist/runtime/query/__tests__/provider-fault-pauses.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/provider-fault-pauses.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/provider-fault-pauses.test.js +86 -0
- package/dist/runtime/query/__tests__/provider-fault-pauses.test.js.map +1 -0
- package/dist/runtime/query/__tests__/reasoning-stream.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/reasoning-stream.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/reasoning-stream.test.js +146 -0
- package/dist/runtime/query/__tests__/reasoning-stream.test.js.map +1 -0
- package/dist/runtime/query/__tests__/recoverable-settle.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/recoverable-settle.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/recoverable-settle.test.js +119 -0
- package/dist/runtime/query/__tests__/recoverable-settle.test.js.map +1 -0
- package/dist/runtime/query/__tests__/resume-budget.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/resume-budget.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/resume-budget.test.js +130 -0
- package/dist/runtime/query/__tests__/resume-budget.test.js.map +1 -0
- package/dist/runtime/query/__tests__/resume-pending.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/resume-pending.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/resume-pending.test.js +226 -0
- package/dist/runtime/query/__tests__/resume-pending.test.js.map +1 -0
- package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.js +89 -0
- package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.js.map +1 -0
- package/dist/runtime/query/__tests__/stop-condition.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/stop-condition.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/stop-condition.test.js +205 -0
- package/dist/runtime/query/__tests__/stop-condition.test.js.map +1 -0
- package/dist/runtime/query/__tests__/structured-output.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/structured-output.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/structured-output.test.js +219 -0
- package/dist/runtime/query/__tests__/structured-output.test.js.map +1 -0
- package/dist/runtime/query/__tests__/structured-run-failure.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/structured-run-failure.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/structured-run-failure.test.js +117 -0
- package/dist/runtime/query/__tests__/structured-run-failure.test.js.map +1 -0
- package/dist/runtime/query/__tests__/terminal-tool.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/terminal-tool.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/terminal-tool.test.js +119 -0
- package/dist/runtime/query/__tests__/terminal-tool.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-grants.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-grants.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-grants.test.js +85 -0
- package/dist/runtime/query/__tests__/tool-grants.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-output-budget.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-output-budget.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-output-budget.test.js +0 -0
- package/dist/runtime/query/__tests__/tool-output-budget.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-pause.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-pause.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-pause.test.js +165 -0
- package/dist/runtime/query/__tests__/tool-pause.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-progress.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-progress.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-progress.test.js +122 -0
- package/dist/runtime/query/__tests__/tool-progress.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-repair-retry.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-repair-retry.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-repair-retry.test.js +196 -0
- package/dist/runtime/query/__tests__/tool-repair-retry.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-result-wire.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-result-wire.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-result-wire.test.js +106 -0
- package/dist/runtime/query/__tests__/tool-result-wire.test.js.map +1 -0
- package/dist/runtime/query/__tests__/tool-timeout.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/tool-timeout.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/tool-timeout.test.js +184 -0
- package/dist/runtime/query/__tests__/tool-timeout.test.js.map +1 -0
- package/dist/runtime/query/__tests__/trace-parenting.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/trace-parenting.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/trace-parenting.test.js +111 -0
- package/dist/runtime/query/__tests__/trace-parenting.test.js.map +1 -0
- package/dist/runtime/query/__tests__/wiring-seams.test.d.ts +2 -0
- package/dist/runtime/query/__tests__/wiring-seams.test.d.ts.map +1 -0
- package/dist/runtime/query/__tests__/wiring-seams.test.js +237 -0
- package/dist/runtime/query/__tests__/wiring-seams.test.js.map +1 -0
- package/dist/runtime/query/checkpoint.d.ts +124 -6
- package/dist/runtime/query/checkpoint.d.ts.map +1 -1
- package/dist/runtime/query/checkpoint.js +207 -6
- package/dist/runtime/query/checkpoint.js.map +1 -1
- package/dist/runtime/query/events.js +2 -2
- package/dist/runtime/query/events.js.map +1 -1
- package/dist/runtime/query/executor.d.ts +224 -7
- package/dist/runtime/query/executor.d.ts.map +1 -1
- package/dist/runtime/query/executor.js +662 -35
- package/dist/runtime/query/executor.js.map +1 -1
- package/dist/runtime/query/guard.d.ts +18 -0
- package/dist/runtime/query/guard.d.ts.map +1 -1
- package/dist/runtime/query/guard.js +16 -1
- package/dist/runtime/query/guard.js.map +1 -1
- package/dist/runtime/query/guardrail-presets.d.ts +36 -0
- package/dist/runtime/query/guardrail-presets.d.ts.map +1 -0
- package/dist/runtime/query/guardrail-presets.js +96 -0
- package/dist/runtime/query/guardrail-presets.js.map +1 -0
- package/dist/runtime/query/guardrails.d.ts +35 -0
- package/dist/runtime/query/guardrails.d.ts.map +1 -0
- package/dist/runtime/query/guardrails.js +102 -0
- package/dist/runtime/query/guardrails.js.map +1 -0
- package/dist/runtime/query/index.d.ts +213 -2
- package/dist/runtime/query/index.d.ts.map +1 -1
- package/dist/runtime/query/index.js +458 -10
- package/dist/runtime/query/index.js.map +1 -1
- package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.d.ts +2 -0
- package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.d.ts.map +1 -0
- package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.js +131 -0
- package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.js.map +1 -0
- package/dist/runtime/query/iteration/index.d.ts +76 -1
- package/dist/runtime/query/iteration/index.d.ts.map +1 -1
- package/dist/runtime/query/iteration/index.js +577 -34
- package/dist/runtime/query/iteration/index.js.map +1 -1
- package/dist/runtime/query/iteration/phases/advisory.d.ts.map +1 -1
- package/dist/runtime/query/iteration/phases/advisory.js +14 -2
- package/dist/runtime/query/iteration/phases/advisory.js.map +1 -1
- package/dist/runtime/query/iteration/phases/advisory.test.d.ts +1 -1
- package/dist/runtime/query/iteration/phases/advisory.test.js +4 -1
- package/dist/runtime/query/iteration/phases/advisory.test.js.map +1 -1
- package/dist/runtime/query/iteration/phases/checkpoint.js +2 -2
- package/dist/runtime/query/iteration/phases/checkpoint.js.map +1 -1
- package/dist/runtime/query/iteration/phases/compaction-safe-cut.test.js +1 -1
- package/dist/runtime/query/iteration/phases/compaction-safe-cut.test.js.map +1 -1
- package/dist/runtime/query/iteration/phases/compaction.d.ts +22 -1
- package/dist/runtime/query/iteration/phases/compaction.d.ts.map +1 -1
- package/dist/runtime/query/iteration/phases/compaction.js +297 -64
- package/dist/runtime/query/iteration/phases/compaction.js.map +1 -1
- package/dist/runtime/query/iteration/phases/context-measurement.test.d.ts +21 -0
- package/dist/runtime/query/iteration/phases/context-measurement.test.d.ts.map +1 -0
- package/dist/runtime/query/iteration/phases/context-measurement.test.js +161 -0
- package/dist/runtime/query/iteration/phases/context-measurement.test.js.map +1 -0
- package/dist/runtime/query/iteration/phases/context.d.ts +62 -2
- package/dist/runtime/query/iteration/phases/context.d.ts.map +1 -1
- package/dist/runtime/query/iteration/phases/context.js +87 -0
- package/dist/runtime/query/iteration/phases/context.js.map +1 -1
- package/dist/runtime/query/iteration/phases/hitl-answer-question.test.js +3 -1
- package/dist/runtime/query/iteration/phases/hitl-answer-question.test.js.map +1 -1
- package/dist/runtime/query/iteration/phases/overflow-escalation.test.d.ts +15 -0
- package/dist/runtime/query/iteration/phases/overflow-escalation.test.d.ts.map +1 -0
- package/dist/runtime/query/iteration/phases/overflow-escalation.test.js +130 -0
- package/dist/runtime/query/iteration/phases/overflow-escalation.test.js.map +1 -0
- package/dist/runtime/query/iteration/phases/plan.d.ts.map +1 -1
- package/dist/runtime/query/iteration/phases/plan.js +8 -2
- package/dist/runtime/query/iteration/phases/plan.js.map +1 -1
- package/dist/runtime/query/iteration/phases/retained-messages.test.d.ts +2 -0
- package/dist/runtime/query/iteration/phases/retained-messages.test.d.ts.map +1 -0
- package/dist/runtime/query/iteration/phases/retained-messages.test.js +180 -0
- package/dist/runtime/query/iteration/phases/retained-messages.test.js.map +1 -0
- package/dist/runtime/query/iteration/phases/tool-review-denials.test.d.ts +2 -0
- package/dist/runtime/query/iteration/phases/tool-review-denials.test.d.ts.map +1 -0
- package/dist/runtime/query/iteration/phases/tool-review-denials.test.js +224 -0
- package/dist/runtime/query/iteration/phases/tool-review-denials.test.js.map +1 -0
- package/dist/runtime/query/iteration/phases/tool-review.d.ts +29 -1
- package/dist/runtime/query/iteration/phases/tool-review.d.ts.map +1 -1
- package/dist/runtime/query/iteration/phases/tool-review.js +114 -44
- package/dist/runtime/query/iteration/phases/tool-review.js.map +1 -1
- package/dist/runtime/query/iteration/phases/working-memory-compaction.test.js +20 -9
- package/dist/runtime/query/iteration/phases/working-memory-compaction.test.js.map +1 -1
- package/dist/runtime/query/iteration/stream-turn.d.ts +2 -26
- package/dist/runtime/query/iteration/stream-turn.d.ts.map +1 -1
- package/dist/runtime/query/iteration/stream-turn.js +254 -16
- package/dist/runtime/query/iteration/stream-turn.js.map +1 -1
- package/dist/runtime/query/plugin-hooks.d.ts.map +1 -1
- package/dist/runtime/query/plugin-hooks.js +6 -2
- package/dist/runtime/query/plugin-hooks.js.map +1 -1
- package/dist/runtime/query/prompt.d.ts.map +1 -1
- package/dist/runtime/query/prompt.js +25 -1
- package/dist/runtime/query/prompt.js.map +1 -1
- package/dist/runtime/query/question-park.d.ts +85 -0
- package/dist/runtime/query/question-park.d.ts.map +1 -0
- package/dist/runtime/query/question-park.js +71 -0
- package/dist/runtime/query/question-park.js.map +1 -0
- package/dist/runtime/query/result.d.ts +9 -0
- package/dist/runtime/query/result.d.ts.map +1 -1
- package/dist/runtime/query/result.js +49 -0
- package/dist/runtime/query/result.js.map +1 -1
- package/dist/runtime/query/resume-pending.d.ts +106 -0
- package/dist/runtime/query/resume-pending.d.ts.map +1 -0
- package/dist/runtime/query/resume-pending.js +312 -0
- package/dist/runtime/query/resume-pending.js.map +1 -0
- package/dist/runtime/query/run-state.d.ts +52 -0
- package/dist/runtime/query/run-state.d.ts.map +1 -0
- package/dist/runtime/query/run-state.js +87 -0
- package/dist/runtime/query/run-state.js.map +1 -0
- package/dist/runtime/query/tool-grants.d.ts +44 -0
- package/dist/runtime/query/tool-grants.d.ts.map +1 -0
- package/dist/runtime/query/tool-grants.js +63 -0
- package/dist/runtime/query/tool-grants.js.map +1 -0
- package/dist/runtime/query/tool-output-budget.d.ts +83 -0
- package/dist/runtime/query/tool-output-budget.d.ts.map +1 -0
- package/dist/runtime/query/tool-output-budget.js +147 -0
- package/dist/runtime/query/tool-output-budget.js.map +1 -0
- package/dist/runtime/query/tool-pause.d.ts +41 -0
- package/dist/runtime/query/tool-pause.d.ts.map +1 -0
- package/dist/runtime/query/tool-pause.js +94 -0
- package/dist/runtime/query/tool-pause.js.map +1 -0
- package/dist/runtime/query/tooling.d.ts +10 -1
- package/dist/runtime/query/tooling.d.ts.map +1 -1
- package/dist/runtime/query/tooling.js +13 -0
- package/dist/runtime/query/tooling.js.map +1 -1
- package/dist/sandbox/__tests__/isolation.test.d.ts +2 -0
- package/dist/sandbox/__tests__/isolation.test.d.ts.map +1 -0
- package/dist/sandbox/__tests__/isolation.test.js +139 -0
- package/dist/sandbox/__tests__/isolation.test.js.map +1 -0
- package/dist/sandbox/__tests__/output-truncation.test.d.ts +2 -0
- package/dist/sandbox/__tests__/output-truncation.test.d.ts.map +1 -0
- package/dist/sandbox/__tests__/output-truncation.test.js +88 -0
- package/dist/sandbox/__tests__/output-truncation.test.js.map +1 -0
- package/dist/sandbox/__tests__/resource-limits.test.d.ts +2 -0
- package/dist/sandbox/__tests__/resource-limits.test.d.ts.map +1 -0
- package/dist/sandbox/__tests__/resource-limits.test.js +86 -0
- package/dist/sandbox/__tests__/resource-limits.test.js.map +1 -0
- package/dist/sandbox/factory.js +1 -1
- package/dist/sandbox/factory.js.map +1 -1
- package/dist/sandbox/index.d.ts +2 -0
- package/dist/sandbox/index.d.ts.map +1 -1
- package/dist/sandbox/index.js +1 -0
- package/dist/sandbox/index.js.map +1 -1
- package/dist/sandbox/isolation.d.ts +15 -0
- package/dist/sandbox/isolation.d.ts.map +1 -0
- package/dist/sandbox/isolation.js +53 -0
- package/dist/sandbox/isolation.js.map +1 -0
- package/dist/sandbox/provider/local.d.ts +51 -2
- package/dist/sandbox/provider/local.d.ts.map +1 -1
- package/dist/sandbox/provider/local.js +139 -69
- package/dist/sandbox/provider/local.js.map +1 -1
- package/dist/session/__tests__/integration/spawn-rollback.test.d.ts +1 -1
- package/dist/session/__tests__/integration/spawn-rollback.test.js +1 -1
- package/dist/session/handoff/broadcast.js +3 -3
- package/dist/session/handoff/broadcast.js.map +1 -1
- package/dist/session/migration/__tests__/filesystem.test.js +5 -1
- package/dist/session/migration/__tests__/filesystem.test.js.map +1 -1
- package/dist/session/migration/filesystem.d.ts.map +1 -1
- package/dist/session/migration/filesystem.js +7 -13
- package/dist/session/migration/filesystem.js.map +1 -1
- package/dist/session/migration/marker.d.ts.map +1 -1
- package/dist/session/migration/marker.js +5 -10
- package/dist/session/migration/marker.js.map +1 -1
- package/dist/session/retention/disk-backend.d.ts.map +1 -1
- package/dist/session/retention/disk-backend.js +5 -12
- package/dist/session/retention/disk-backend.js.map +1 -1
- package/dist/session/workspace/__tests__/git-worktree.test.js +6 -2
- package/dist/session/workspace/__tests__/git-worktree.test.js.map +1 -1
- package/dist/session/workspace/__tests__/shared-run.test.js +15 -0
- package/dist/session/workspace/__tests__/shared-run.test.js.map +1 -1
- package/dist/session/workspace/shared-run.js +8 -1
- package/dist/session/workspace/shared-run.js.map +1 -1
- package/dist/skills/__tests__/frontmatter.test.d.ts +2 -0
- package/dist/skills/__tests__/frontmatter.test.d.ts.map +1 -0
- package/dist/skills/__tests__/frontmatter.test.js +106 -0
- package/dist/skills/__tests__/frontmatter.test.js.map +1 -0
- package/dist/skills/loader.d.ts.map +1 -1
- package/dist/skills/loader.js +42 -3
- package/dist/skills/loader.js.map +1 -1
- package/dist/store/__tests__/checkpoint-integrity.test.d.ts +2 -0
- package/dist/store/__tests__/checkpoint-integrity.test.d.ts.map +1 -0
- package/dist/store/__tests__/checkpoint-integrity.test.js +112 -0
- package/dist/store/__tests__/checkpoint-integrity.test.js.map +1 -0
- package/dist/store/__tests__/checkpoint-versioning.test.d.ts +2 -0
- package/dist/store/__tests__/checkpoint-versioning.test.d.ts.map +1 -0
- package/dist/store/__tests__/checkpoint-versioning.test.js +151 -0
- package/dist/store/__tests__/checkpoint-versioning.test.js.map +1 -0
- package/dist/store/__tests__/schema-adoption.test.d.ts +2 -0
- package/dist/store/__tests__/schema-adoption.test.d.ts.map +1 -0
- package/dist/store/__tests__/schema-adoption.test.js +90 -0
- package/dist/store/__tests__/schema-adoption.test.js.map +1 -0
- package/dist/store/__tests__/schema.test.d.ts +2 -0
- package/dist/store/__tests__/schema.test.d.ts.map +1 -0
- package/dist/store/__tests__/schema.test.js +93 -0
- package/dist/store/__tests__/schema.test.js.map +1 -0
- package/dist/store/memory/disk.d.ts.map +1 -1
- package/dist/store/memory/disk.js +16 -16
- package/dist/store/memory/disk.js.map +1 -1
- package/dist/store/run/disk.d.ts +23 -0
- package/dist/store/run/disk.d.ts.map +1 -1
- package/dist/store/run/disk.js +148 -23
- package/dist/store/run/disk.js.map +1 -1
- package/dist/store/schema.d.ts +73 -0
- package/dist/store/schema.d.ts.map +1 -0
- package/dist/store/schema.js +103 -0
- package/dist/store/schema.js.map +1 -0
- package/dist/store/session/disk.d.ts.map +1 -1
- package/dist/store/session/disk.js +25 -13
- package/dist/store/session/disk.js.map +1 -1
- package/dist/store/task/__tests__/lookup-across-runs.test.d.ts +2 -0
- package/dist/store/task/__tests__/lookup-across-runs.test.d.ts.map +1 -0
- package/dist/store/task/__tests__/lookup-across-runs.test.js +60 -0
- package/dist/store/task/__tests__/lookup-across-runs.test.js.map +1 -0
- package/dist/store/task/disk.d.ts +18 -0
- package/dist/store/task/disk.d.ts.map +1 -1
- package/dist/store/task/disk.js +61 -16
- package/dist/store/task/disk.js.map +1 -1
- package/dist/store/thread/disk.d.ts.map +1 -1
- package/dist/store/thread/disk.js +20 -11
- package/dist/store/thread/disk.js.map +1 -1
- package/dist/streaming/coalesce.d.ts +1 -1
- package/dist/streaming/coalesce.js +1 -1
- package/dist/telemetry/__tests__/latency-instruments.test.d.ts +2 -0
- package/dist/telemetry/__tests__/latency-instruments.test.d.ts.map +1 -0
- package/dist/telemetry/__tests__/latency-instruments.test.js +82 -0
- package/dist/telemetry/__tests__/latency-instruments.test.js.map +1 -0
- package/dist/telemetry/__tests__/metrics.test.d.ts +2 -0
- package/dist/telemetry/__tests__/metrics.test.d.ts.map +1 -0
- package/dist/telemetry/__tests__/metrics.test.js +112 -0
- package/dist/telemetry/__tests__/metrics.test.js.map +1 -0
- package/dist/telemetry/__tests__/span-hierarchy.test.d.ts +2 -0
- package/dist/telemetry/__tests__/span-hierarchy.test.d.ts.map +1 -0
- package/dist/telemetry/__tests__/span-hierarchy.test.js +84 -0
- package/dist/telemetry/__tests__/span-hierarchy.test.js.map +1 -0
- package/dist/telemetry/attributes.d.ts +44 -0
- package/dist/telemetry/attributes.d.ts.map +1 -1
- package/dist/telemetry/attributes.js +80 -0
- package/dist/telemetry/attributes.js.map +1 -1
- package/dist/telemetry/metrics.d.ts +41 -0
- package/dist/telemetry/metrics.d.ts.map +1 -0
- package/dist/telemetry/metrics.js +160 -0
- package/dist/telemetry/metrics.js.map +1 -0
- package/dist/test-setup.d.ts +3 -3
- package/dist/test-setup.js +3 -3
- package/dist/test-support/paths.d.ts +14 -0
- package/dist/test-support/paths.d.ts.map +1 -0
- package/dist/test-support/paths.js +16 -0
- package/dist/test-support/paths.js.map +1 -0
- package/dist/tools/__tests__/path-containment.test.d.ts +2 -0
- package/dist/tools/__tests__/path-containment.test.d.ts.map +1 -0
- package/dist/tools/__tests__/path-containment.test.js +133 -0
- package/dist/tools/__tests__/path-containment.test.js.map +1 -0
- package/dist/tools/__tests__/posix-path.test.d.ts +2 -0
- package/dist/tools/__tests__/posix-path.test.d.ts.map +1 -0
- package/dist/tools/__tests__/posix-path.test.js +67 -0
- package/dist/tools/__tests__/posix-path.test.js.map +1 -0
- package/dist/tools/__tests__/sandboxed-search.test.d.ts +2 -0
- package/dist/tools/__tests__/sandboxed-search.test.d.ts.map +1 -0
- package/dist/tools/__tests__/sandboxed-search.test.js +114 -0
- package/dist/tools/__tests__/sandboxed-search.test.js.map +1 -0
- package/dist/tools/builtins/__tests__/computer-use.test.js +12 -2
- package/dist/tools/builtins/__tests__/computer-use.test.js.map +1 -1
- package/dist/tools/builtins/__tests__/edit.test.js +10 -216
- package/dist/tools/builtins/__tests__/edit.test.js.map +1 -1
- package/dist/tools/builtins/__tests__/payload-budget.test.js +16 -30
- package/dist/tools/builtins/__tests__/payload-budget.test.js.map +1 -1
- package/dist/tools/builtins/__tests__/read-file.test.js +23 -1
- package/dist/tools/builtins/__tests__/read-file.test.js.map +1 -1
- package/dist/tools/builtins/__tests__/write-file.test.js +13 -37
- package/dist/tools/builtins/__tests__/write-file.test.js.map +1 -1
- package/dist/tools/builtins/bash.d.ts.map +1 -1
- package/dist/tools/builtins/bash.js +33 -4
- package/dist/tools/builtins/bash.js.map +1 -1
- package/dist/tools/builtins/computer-use.d.ts +22 -22
- package/dist/tools/builtins/computer-use.d.ts.map +1 -1
- package/dist/tools/builtins/computer-use.js +7 -1
- package/dist/tools/builtins/computer-use.js.map +1 -1
- package/dist/tools/builtins/edit.d.ts +5 -2
- package/dist/tools/builtins/edit.d.ts.map +1 -1
- package/dist/tools/builtins/edit.js +125 -105
- package/dist/tools/builtins/edit.js.map +1 -1
- package/dist/tools/builtins/glob.d.ts +1 -1
- package/dist/tools/builtins/glob.d.ts.map +1 -1
- package/dist/tools/builtins/glob.js +63 -6
- package/dist/tools/builtins/glob.js.map +1 -1
- package/dist/tools/builtins/grep.d.ts.map +1 -1
- package/dist/tools/builtins/grep.js +61 -8
- package/dist/tools/builtins/grep.js.map +1 -1
- package/dist/tools/builtins/index.js +6 -6
- package/dist/tools/builtins/index.js.map +1 -1
- package/dist/tools/builtins/ls.d.ts.map +1 -1
- package/dist/tools/builtins/ls.js +4 -2
- package/dist/tools/builtins/ls.js.map +1 -1
- package/dist/tools/builtins/read-file.d.ts +1 -1
- package/dist/tools/builtins/read-file.d.ts.map +1 -1
- package/dist/tools/builtins/read-file.js +38 -4
- package/dist/tools/builtins/read-file.js.map +1 -1
- package/dist/tools/builtins/write-file.d.ts +2 -1
- package/dist/tools/builtins/write-file.d.ts.map +1 -1
- package/dist/tools/builtins/write-file.js +40 -66
- package/dist/tools/builtins/write-file.js.map +1 -1
- package/dist/tools/coordinator/__tests__/agent.test.d.ts +1 -1
- package/dist/tools/coordinator/__tests__/agent.test.js +2 -2
- package/dist/tools/coordinator/__tests__/agent.test.js.map +1 -1
- package/dist/tools/coordinator/agent.d.ts +20 -9
- package/dist/tools/coordinator/agent.d.ts.map +1 -1
- package/dist/tools/coordinator/agent.js +2 -1
- package/dist/tools/coordinator/agent.js.map +1 -1
- package/dist/tools/coordinator/index.d.ts +20 -0
- package/dist/tools/coordinator/index.d.ts.map +1 -1
- package/dist/tools/coordinator/index.js +80 -32
- package/dist/tools/coordinator/index.js.map +1 -1
- package/dist/tools/defineTool.d.ts +16 -0
- package/dist/tools/defineTool.d.ts.map +1 -1
- package/dist/tools/defineTool.js +4 -0
- package/dist/tools/defineTool.js.map +1 -1
- package/dist/tools/glob-match.d.ts +18 -0
- package/dist/tools/glob-match.d.ts.map +1 -0
- package/dist/tools/glob-match.js +50 -0
- package/dist/tools/glob-match.js.map +1 -0
- package/dist/tools/paths.d.ts +20 -0
- package/dist/tools/paths.d.ts.map +1 -0
- package/dist/tools/paths.js +41 -0
- package/dist/tools/paths.js.map +1 -0
- package/dist/tools/posix-path.d.ts +24 -0
- package/dist/tools/posix-path.d.ts.map +1 -0
- package/dist/tools/posix-path.js +67 -0
- package/dist/tools/posix-path.js.map +1 -0
- package/dist/types/advisory/config.d.ts +21 -3
- package/dist/types/advisory/config.d.ts.map +1 -1
- package/dist/types/advisory/result.d.ts +12 -10
- package/dist/types/advisory/result.d.ts.map +1 -1
- package/dist/types/agent/base.d.ts +20 -2
- package/dist/types/agent/base.d.ts.map +1 -1
- package/dist/types/agent/factory.d.ts +15 -9
- package/dist/types/agent/factory.d.ts.map +1 -1
- package/dist/types/agent/gateway.d.ts +18 -0
- package/dist/types/agent/gateway.d.ts.map +1 -1
- package/dist/types/agent/manager.d.ts +19 -0
- package/dist/types/agent/manager.d.ts.map +1 -1
- package/dist/types/agent/reactive.d.ts +48 -0
- package/dist/types/agent/reactive.d.ts.map +1 -1
- package/dist/types/agent/router.d.ts +11 -0
- package/dist/types/agent/router.d.ts.map +1 -1
- package/dist/types/agent/supervisor.d.ts +2 -2
- package/dist/types/agent/task.d.ts +24 -0
- package/dist/types/agent/task.d.ts.map +1 -1
- package/dist/types/bidi/index.d.ts +140 -0
- package/dist/types/bidi/index.d.ts.map +1 -0
- package/dist/types/bidi/index.js +2 -0
- package/dist/types/bidi/index.js.map +1 -0
- package/dist/types/common/__tests__/numeric-correctness.test.d.ts +2 -0
- package/dist/types/common/__tests__/numeric-correctness.test.d.ts.map +1 -0
- package/dist/types/common/__tests__/numeric-correctness.test.js +108 -0
- package/dist/types/common/__tests__/numeric-correctness.test.js.map +1 -0
- package/dist/types/common/index.d.ts +1 -1
- package/dist/types/common/index.d.ts.map +1 -1
- package/dist/types/common/index.js +15 -4
- package/dist/types/common/index.js.map +1 -1
- package/dist/types/connector/mcp.d.ts +29 -0
- package/dist/types/connector/mcp.d.ts.map +1 -1
- package/dist/types/errors/__tests__/catalog.test.d.ts +2 -0
- package/dist/types/errors/__tests__/catalog.test.d.ts.map +1 -0
- package/dist/types/errors/__tests__/catalog.test.js +106 -0
- package/dist/types/errors/__tests__/catalog.test.js.map +1 -0
- package/dist/types/errors/__tests__/errors.test.d.ts +2 -0
- package/dist/types/errors/__tests__/errors.test.d.ts.map +1 -0
- package/dist/types/errors/__tests__/errors.test.js +104 -0
- package/dist/types/errors/__tests__/errors.test.js.map +1 -0
- package/dist/types/errors/catalog.d.ts +83 -0
- package/dist/types/errors/catalog.d.ts.map +1 -0
- package/dist/types/errors/catalog.js +148 -0
- package/dist/types/errors/catalog.js.map +1 -0
- package/dist/types/errors/index.d.ts +73 -0
- package/dist/types/errors/index.d.ts.map +1 -0
- package/dist/types/errors/index.js +70 -0
- package/dist/types/errors/index.js.map +1 -0
- package/dist/types/guardrail/index.d.ts +57 -0
- package/dist/types/guardrail/index.d.ts.map +1 -0
- package/dist/types/guardrail/index.js +2 -0
- package/dist/types/guardrail/index.js.map +1 -0
- package/dist/types/hitl/index.d.ts +97 -14
- package/dist/types/hitl/index.d.ts.map +1 -1
- package/dist/types/hitl/index.js.map +1 -1
- package/dist/types/ids/index.d.ts +2 -2
- package/dist/types/message/content.d.ts +22 -0
- package/dist/types/message/content.d.ts.map +1 -0
- package/dist/types/message/content.js +49 -0
- package/dist/types/message/content.js.map +1 -0
- package/dist/types/message/index.d.ts +193 -8
- package/dist/types/message/index.d.ts.map +1 -1
- package/dist/types/message/index.js +19 -4
- package/dist/types/message/index.js.map +1 -1
- package/dist/types/plugin/index.d.ts +78 -3
- package/dist/types/plugin/index.d.ts.map +1 -1
- package/dist/types/plugin/index.js +0 -1
- package/dist/types/plugin/index.js.map +1 -1
- package/dist/types/provider/__tests__/cause-chain.test.d.ts +2 -0
- package/dist/types/provider/__tests__/cause-chain.test.d.ts.map +1 -0
- package/dist/types/provider/__tests__/cause-chain.test.js +108 -0
- package/dist/types/provider/__tests__/cause-chain.test.js.map +1 -0
- package/dist/types/provider/__tests__/declared-retryable.test.d.ts +2 -0
- package/dist/types/provider/__tests__/declared-retryable.test.d.ts.map +1 -0
- package/dist/types/provider/__tests__/declared-retryable.test.js +70 -0
- package/dist/types/provider/__tests__/declared-retryable.test.js.map +1 -0
- package/dist/types/provider/__tests__/overflow-classification.test.d.ts +2 -0
- package/dist/types/provider/__tests__/overflow-classification.test.d.ts.map +1 -0
- package/dist/types/provider/__tests__/overflow-classification.test.js +80 -0
- package/dist/types/provider/__tests__/overflow-classification.test.js.map +1 -0
- package/dist/types/provider/chat.d.ts +27 -1
- package/dist/types/provider/chat.d.ts.map +1 -1
- package/dist/types/provider/config.d.ts +83 -2
- package/dist/types/provider/config.d.ts.map +1 -1
- package/dist/types/provider/errors.d.ts +102 -0
- package/dist/types/provider/errors.d.ts.map +1 -0
- package/dist/types/provider/errors.js +393 -0
- package/dist/types/provider/errors.js.map +1 -0
- package/dist/types/provider/index.d.ts +3 -2
- package/dist/types/provider/index.d.ts.map +1 -1
- package/dist/types/provider/stream.d.ts +59 -3
- package/dist/types/provider/stream.d.ts.map +1 -1
- package/dist/types/rag/embedding.d.ts +11 -2
- package/dist/types/rag/embedding.d.ts.map +1 -1
- package/dist/types/run/__tests__/step.test.d.ts +2 -0
- package/dist/types/run/__tests__/step.test.d.ts.map +1 -0
- package/dist/types/run/__tests__/step.test.js +107 -0
- package/dist/types/run/__tests__/step.test.js.map +1 -0
- package/dist/types/run/answer-review.d.ts +46 -0
- package/dist/types/run/answer-review.d.ts.map +1 -0
- package/dist/types/run/answer-review.js +2 -0
- package/dist/types/run/answer-review.js.map +1 -0
- package/dist/types/run/config.d.ts +33 -0
- package/dist/types/run/config.d.ts.map +1 -1
- package/dist/types/run/derive-status.d.ts +29 -0
- package/dist/types/run/derive-status.d.ts.map +1 -0
- package/dist/types/run/derive-status.js +49 -0
- package/dist/types/run/derive-status.js.map +1 -0
- package/dist/types/run/entity.d.ts +16 -0
- package/dist/types/run/entity.d.ts.map +1 -1
- package/dist/types/run/events.d.ts +228 -4
- package/dist/types/run/events.d.ts.map +1 -1
- package/dist/types/run/events.js +7 -0
- package/dist/types/run/events.js.map +1 -1
- package/dist/types/run/index.d.ts +5 -0
- package/dist/types/run/index.d.ts.map +1 -1
- package/dist/types/run/index.js +3 -0
- package/dist/types/run/index.js.map +1 -1
- package/dist/types/run/memory-promotion.d.ts +90 -0
- package/dist/types/run/memory-promotion.d.ts.map +1 -0
- package/dist/types/run/memory-promotion.js +38 -0
- package/dist/types/run/memory-promotion.js.map +1 -0
- package/dist/types/run/prepare-step.d.ts +100 -0
- package/dist/types/run/prepare-step.d.ts.map +1 -0
- package/dist/types/run/prepare-step.js +2 -0
- package/dist/types/run/prepare-step.js.map +1 -0
- package/dist/types/run/state.d.ts +85 -0
- package/dist/types/run/state.d.ts.map +1 -0
- package/dist/types/run/state.js +33 -0
- package/dist/types/run/state.js.map +1 -0
- package/dist/types/run/step.d.ts +77 -0
- package/dist/types/run/step.d.ts.map +1 -0
- package/dist/types/run/step.js +26 -0
- package/dist/types/run/step.js.map +1 -0
- package/dist/types/run/stop-reason.d.ts +31 -3
- package/dist/types/run/stop-reason.d.ts.map +1 -1
- package/dist/types/sandbox/index.d.ts +82 -10
- package/dist/types/sandbox/index.d.ts.map +1 -1
- package/dist/types/sandbox/index.js +13 -1
- package/dist/types/sandbox/index.js.map +1 -1
- package/dist/types/structured-output/index.d.ts +6 -5
- package/dist/types/structured-output/index.d.ts.map +1 -1
- package/dist/types/thread/entity.d.ts +2 -3
- package/dist/types/thread/entity.d.ts.map +1 -1
- package/dist/types/tool/index.d.ts +193 -1
- package/dist/types/tool/index.d.ts.map +1 -1
- package/dist/types/tool/index.js +1 -1
- package/dist/types/tool/index.js.map +1 -1
- package/dist/types/tool/repair.d.ts +49 -0
- package/dist/types/tool/repair.d.ts.map +1 -0
- package/dist/types/tool/repair.js +2 -0
- package/dist/types/tool/repair.js.map +1 -0
- package/dist/types/toolset/index.d.ts +2 -5
- package/dist/types/toolset/index.d.ts.map +1 -1
- package/dist/utils/__tests__/atomic-write.test.d.ts +2 -0
- package/dist/utils/__tests__/atomic-write.test.d.ts.map +1 -0
- package/dist/utils/__tests__/atomic-write.test.js +91 -0
- package/dist/utils/__tests__/atomic-write.test.js.map +1 -0
- package/dist/utils/__tests__/shell-compress.test.d.ts +2 -0
- package/dist/utils/__tests__/shell-compress.test.d.ts.map +1 -0
- package/dist/utils/__tests__/shell-compress.test.js +41 -0
- package/dist/utils/__tests__/shell-compress.test.js.map +1 -0
- package/dist/utils/atomic-write.d.ts +10 -0
- package/dist/utils/atomic-write.d.ts.map +1 -0
- package/dist/utils/atomic-write.js +83 -0
- package/dist/utils/atomic-write.js.map +1 -0
- package/dist/utils/shell-compress.d.ts.map +1 -1
- package/dist/utils/shell-compress.js +12 -1
- package/dist/utils/shell-compress.js.map +1 -1
- package/dist/verification/presets.d.ts +6 -3
- package/dist/verification/presets.d.ts.map +1 -1
- package/dist/verification/presets.js +6 -3
- package/dist/verification/presets.js.map +1 -1
- package/package.json +3 -3
- package/src/__tests__/default-path-conformance.test.ts +306 -0
- package/src/__tests__/shipped-defaults.test.ts +111 -0
- package/src/advisory/__tests__/budget-enforcement.test.ts +208 -0
- package/src/advisory/__tests__/structured-response.test.ts +128 -0
- package/src/advisory/budget.ts +27 -0
- package/src/advisory/context.ts +19 -0
- package/src/advisory/executor.test.ts +12 -8
- package/src/advisory/executor.ts +37 -25
- package/src/advisory/index.ts +2 -0
- package/src/advisory/parse.ts +63 -0
- package/src/agents/AbstractAgent.ts +82 -0
- package/src/agents/PipelineAgent.ts +19 -0
- package/src/agents/ReactiveAgent.ts +47 -0
- package/src/agents/RouterAgent.ts +32 -2
- package/src/agents/SupervisorAgent.ts +31 -1
- package/src/agents/__tests__/idempotency-key.test.ts +177 -0
- package/src/agents/__tests__/invocation-lock.test.ts +154 -0
- package/src/agents/__tests__/reachability.test.ts +224 -0
- package/src/agents/__tests__/supervisor-ledger.test.ts +1 -1
- package/src/agents/index.ts +1 -1
- package/src/agents/lock.ts +0 -2
- package/src/bridge/a2a/late-events.test.ts +84 -0
- package/src/bridge/a2a/mapper.test.ts +1 -1
- package/src/bridge/a2a/mapper.ts +70 -5
- package/src/bridge/a2a/message.ts +5 -1
- package/src/bridge/sse/late-events.test.ts +182 -0
- package/src/bridge/sse/mapper.ts +109 -0
- package/src/compaction/__tests__/cleared-result-recovery.test.ts +118 -0
- package/src/compaction/__tests__/context-window.test.ts +77 -0
- package/src/compaction/__tests__/dangling.test.ts +16 -6
- package/src/compaction/__tests__/lossiness.test.ts +187 -0
- package/src/compaction/__tests__/state-across-resume.test.ts +99 -0
- package/src/compaction/__tests__/tool-result-editing.test.ts +213 -0
- package/src/compaction/__tests__/trim-opens-on-user.test.ts +118 -0
- package/src/compaction/__tests__/verifier-model.test.ts +83 -0
- package/src/compaction/context-window.ts +107 -0
- package/src/compaction/dangling.ts +48 -1
- package/src/compaction/extractor.ts +56 -1
- package/src/compaction/index.ts +16 -0
- package/src/compaction/manager.ts +53 -8
- package/src/compaction/managers/structured.ts +7 -6
- package/src/compaction/retention.ts +89 -0
- package/src/compaction/serializer.ts +42 -11
- package/src/compaction/tool-result-editing.ts +229 -0
- package/src/compaction/types.ts +10 -0
- package/src/compaction/verifier.ts +24 -1
- package/src/compaction/wire.ts +91 -0
- package/src/config/runtime.ts +63 -19
- package/src/connector/builtins/http.test.ts +13 -0
- package/src/connector/builtins/http.ts +15 -1
- package/src/connector/mcp/__tests__/discovery-policy.test.ts +133 -0
- package/src/connector/mcp/__tests__/list-pagination.test.ts +98 -0
- package/src/connector/mcp/__tests__/policy.test.ts +100 -0
- package/src/connector/mcp/__tests__/protocol-negotiation.test.ts +105 -0
- package/src/connector/mcp/__tests__/schema-fidelity.test.ts +250 -0
- package/src/connector/mcp/__tests__/structured-result.test.ts +143 -0
- package/src/connector/mcp/adapter.ts +306 -19
- package/src/connector/mcp/client.test.ts +159 -0
- package/src/connector/mcp/client.ts +160 -13
- package/src/connector/mcp/discovery.ts +74 -5
- package/src/connector/mcp/http-sse.ts +16 -1
- package/src/connector/mcp/index.ts +4 -0
- package/src/connector/mcp/policy.ts +134 -0
- package/src/connector/mcp/schema-refs.ts +126 -0
- package/src/constants/agent/index.ts +3 -0
- package/src/constants/compaction/index.ts +1 -1
- package/src/constants/continuation.ts +4 -4
- package/src/constants/mcp/index.ts +34 -0
- package/src/constants/plugin/index.ts +24 -2
- package/src/constants/sandbox/index.ts +6 -5
- package/src/constants/telemetry/index.ts +8 -0
- package/src/constants/tools/index.ts +20 -0
- package/src/contracts/api.ts +40 -0
- package/src/contracts/run-status.ts +43 -0
- package/src/eval/__tests__/case-deadline.test.ts +147 -0
- package/src/eval/__tests__/end-to-end.test.ts +259 -0
- package/src/eval/__tests__/harness-honesty.test.ts +99 -0
- package/src/eval/__tests__/harness.test.ts +249 -0
- package/src/eval/__tests__/judge.test.ts +306 -0
- package/src/eval/__tests__/scorer-gates.test.ts +156 -0
- package/src/eval/experiment.ts +335 -0
- package/src/eval/from-run.ts +48 -0
- package/src/eval/index.ts +21 -0
- package/src/eval/judge.ts +240 -0
- package/src/eval/scorers.ts +169 -0
- package/src/eval/types.ts +157 -0
- package/src/execution/local.test.ts +46 -0
- package/src/execution/local.ts +6 -1
- package/src/gateway/__tests__/list-survives-eviction.test.ts +124 -0
- package/src/gateway/local.ts +117 -3
- package/src/manager/agent/__tests__/dispose-cancels.test.ts +104 -0
- package/src/manager/agent/__tests__/lifecycle.test.ts +297 -0
- package/src/manager/agent/lifecycle.ts +127 -7
- package/src/manager/run/persistence.ts +132 -1
- package/src/plugin/__tests__/enable-contributions.test.ts +3 -3
- package/src/plugin/__tests__/hook-ordering.test.ts +252 -0
- package/src/plugin/__tests__/install-time-refusal.test.ts +68 -0
- package/src/plugin/__tests__/lifecycle.test.ts +95 -73
- package/src/plugin/lifecycle.ts +76 -22
- package/src/plugin/loader.ts +37 -1
- package/src/probe/registry.test.ts +49 -6
- package/src/probe/registry.ts +21 -1
- package/src/provider/__tests__/mock-script.test.ts +179 -0
- package/src/provider/__tests__/registry.test.ts +1 -1
- package/src/provider/__tests__/retry-visibility.test.ts +160 -0
- package/src/provider/__tests__/retry.test.ts +296 -0
- package/src/provider/capabilities.ts +4 -0
- package/src/provider/collect.test.ts +1 -1
- package/src/provider/collect.ts +1 -1
- package/src/provider/index.ts +8 -0
- package/src/provider/mock-register.ts +5 -2
- package/src/provider/mock.ts +117 -34
- package/src/provider/retry.ts +199 -0
- package/src/public-runtime.ts +111 -7
- package/src/public-types.ts +5 -1
- package/src/rag/__tests__/retrieval-math.test.ts +54 -0
- package/src/rag/chunking.ts +32 -1
- package/src/rag/embedding.test.ts +65 -10
- package/src/rag/embedding.ts +26 -5
- package/src/rag/index.ts +1 -1
- package/src/rag/retriever.ts +92 -21
- package/src/rag/vector-store.test.ts +2 -2
- package/src/registry/ManagedRegistry.test.ts +1 -1
- package/src/registry/Registry.test.ts +1 -1
- package/src/registry/connector/definitions.test.ts +1 -1
- package/src/registry/tool/__tests__/schema.test.ts +208 -0
- package/src/registry/tool/__tests__/tool-name-validation.test.ts +88 -0
- package/src/registry/tool/execute.ts +104 -22
- package/src/registry/tool/schema.ts +80 -0
- package/src/registry/toolset/catalog.test.ts +0 -1
- package/src/registry/toolset/catalog.ts +0 -1
- package/src/run/reporter.ts +84 -0
- package/src/runtime/bidi/__tests__/session.test.ts +232 -0
- package/src/runtime/bidi/index.ts +4 -0
- package/src/runtime/bidi/mock.ts +90 -0
- package/src/runtime/bidi/session.ts +205 -0
- package/src/runtime/decision/parser.test.ts +46 -0
- package/src/runtime/decision/parser.ts +8 -1
- package/src/runtime/query/__tests__/answer-review.test.ts +172 -0
- package/src/runtime/query/__tests__/capability-negotiation.test.ts +25 -129
- package/src/runtime/query/__tests__/checkpoint-store.test.ts +13 -50
- package/src/runtime/query/__tests__/checkpoint.test.ts +2 -10
- package/src/runtime/query/__tests__/citations.test.ts +111 -0
- package/src/runtime/query/__tests__/content-budget.test.ts +83 -0
- package/src/runtime/query/__tests__/context.test.ts +3 -2
- package/src/runtime/query/__tests__/deferred-tools.test.ts +23 -148
- package/src/runtime/query/__tests__/delegation-resume.test.ts +159 -0
- package/src/runtime/query/__tests__/document-attachments.test.ts +125 -0
- package/src/runtime/query/__tests__/durable-park-and-trace.test.ts +245 -0
- package/src/runtime/query/__tests__/durable-question-park.test.ts +305 -0
- package/src/runtime/query/__tests__/durable-run-state.test.ts +194 -0
- package/src/runtime/query/__tests__/executor-plugin-hooks.test.ts +24 -20
- package/src/runtime/query/__tests__/finish-reason-vs-tool-calls.test.ts +206 -0
- package/src/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.ts +115 -0
- package/src/runtime/query/__tests__/guardrails-e2e.test.ts +164 -0
- package/src/runtime/query/__tests__/guardrails.test.ts +236 -0
- package/src/runtime/query/__tests__/hook-model-visibility.test.ts +143 -0
- package/src/runtime/query/__tests__/long-document-flow.test.ts +9 -2
- package/src/runtime/query/__tests__/memory-promotion.test.ts +148 -0
- package/src/runtime/query/__tests__/orphan-task-turn-end.test.ts +5 -31
- package/src/runtime/query/__tests__/overflow-relief.test.ts +139 -0
- package/src/runtime/query/__tests__/partial-batch-resume.test.ts +264 -0
- package/src/runtime/query/__tests__/plugin-hooks.test.ts +0 -1
- package/src/runtime/query/__tests__/prepare-step-chain.test.ts +127 -0
- package/src/runtime/query/__tests__/prepare-step.test.ts +233 -0
- package/src/runtime/query/__tests__/probe-veto-is-error.test.ts +145 -0
- package/src/runtime/query/__tests__/prompt.test.ts +1 -1
- package/src/runtime/query/__tests__/provider-fault-pauses.test.ts +104 -0
- package/src/runtime/query/__tests__/reasoning-stream.test.ts +177 -0
- package/src/runtime/query/__tests__/recoverable-settle.test.ts +145 -0
- package/src/runtime/query/__tests__/resume-budget.test.ts +158 -0
- package/src/runtime/query/__tests__/resume-pending.test.ts +273 -0
- package/src/runtime/query/__tests__/rich-content-budget-wiring.test.ts +109 -0
- package/src/runtime/query/__tests__/stop-condition.test.ts +265 -0
- package/src/runtime/query/__tests__/structured-output.test.ts +268 -0
- package/src/runtime/query/__tests__/structured-run-failure.test.ts +144 -0
- package/src/runtime/query/__tests__/terminal-tool.test.ts +148 -0
- package/src/runtime/query/__tests__/tool-grants.test.ts +102 -0
- package/src/runtime/query/__tests__/tool-output-budget.test.ts +0 -0
- package/src/runtime/query/__tests__/tool-pause.test.ts +209 -0
- package/src/runtime/query/__tests__/tool-progress.test.ts +154 -0
- package/src/runtime/query/__tests__/tool-repair-retry.test.ts +255 -0
- package/src/runtime/query/__tests__/tool-result-wire.test.ts +135 -0
- package/src/runtime/query/__tests__/tool-timeout.test.ts +230 -0
- package/src/runtime/query/__tests__/trace-parenting.test.ts +128 -0
- package/src/runtime/query/__tests__/wiring-seams.test.ts +288 -0
- package/src/runtime/query/checkpoint.ts +242 -10
- package/src/runtime/query/events.ts +2 -2
- package/src/runtime/query/executor.ts +847 -39
- package/src/runtime/query/guard.ts +26 -1
- package/src/runtime/query/guardrail-presets.ts +124 -0
- package/src/runtime/query/guardrails.ts +144 -0
- package/src/runtime/query/index.ts +759 -12
- package/src/runtime/query/iteration/__tests__/cancel-settles-turn.test.ts +177 -0
- package/src/runtime/query/iteration/index.ts +656 -40
- package/src/runtime/query/iteration/phases/advisory.test.ts +4 -1
- package/src/runtime/query/iteration/phases/advisory.ts +14 -2
- package/src/runtime/query/iteration/phases/checkpoint.ts +2 -2
- package/src/runtime/query/iteration/phases/compaction-safe-cut.test.ts +3 -1
- package/src/runtime/query/iteration/phases/compaction.ts +329 -64
- package/src/runtime/query/iteration/phases/context-measurement.test.ts +196 -0
- package/src/runtime/query/iteration/phases/context.ts +155 -2
- package/src/runtime/query/iteration/phases/hitl-answer-question.test.ts +3 -1
- package/src/runtime/query/iteration/phases/overflow-escalation.test.ts +167 -0
- package/src/runtime/query/iteration/phases/plan.ts +10 -2
- package/src/runtime/query/iteration/phases/retained-messages.test.ts +226 -0
- package/src/runtime/query/iteration/phases/tool-review-denials.test.ts +270 -0
- package/src/runtime/query/iteration/phases/tool-review.ts +140 -51
- package/src/runtime/query/iteration/phases/working-memory-compaction.test.ts +21 -9
- package/src/runtime/query/iteration/stream-turn.ts +285 -15
- package/src/runtime/query/plugin-hooks.ts +6 -2
- package/src/runtime/query/prompt.ts +26 -1
- package/src/runtime/query/question-park.ts +120 -0
- package/src/runtime/query/result.ts +64 -0
- package/src/runtime/query/resume-pending.ts +394 -0
- package/src/runtime/query/run-state.ts +124 -0
- package/src/runtime/query/tool-grants.ts +83 -0
- package/src/runtime/query/tool-output-budget.ts +183 -0
- package/src/runtime/query/tool-pause.ts +119 -0
- package/src/runtime/query/tooling.ts +23 -1
- package/src/sandbox/__tests__/isolation.test.ts +164 -0
- package/src/sandbox/__tests__/output-truncation.test.ts +106 -0
- package/src/sandbox/__tests__/resource-limits.test.ts +106 -0
- package/src/sandbox/factory.ts +1 -1
- package/src/sandbox/index.ts +2 -0
- package/src/sandbox/isolation.ts +72 -0
- package/src/sandbox/provider/local.ts +173 -74
- package/src/session/__tests__/integration/spawn-rollback.test.ts +1 -1
- package/src/session/handoff/broadcast.ts +3 -3
- package/src/session/migration/__tests__/filesystem.test.ts +26 -19
- package/src/session/migration/filesystem.ts +7 -12
- package/src/session/migration/marker.ts +5 -9
- package/src/session/retention/disk-backend.ts +5 -11
- package/src/session/workspace/__tests__/git-worktree.test.ts +6 -2
- package/src/session/workspace/__tests__/shared-run.test.ts +18 -0
- package/src/session/workspace/shared-run.ts +7 -1
- package/src/skills/__tests__/frontmatter.test.ts +135 -0
- package/src/skills/loader.ts +47 -3
- package/src/store/__tests__/checkpoint-integrity.test.ts +134 -0
- package/src/store/__tests__/checkpoint-versioning.test.ts +194 -0
- package/src/store/__tests__/schema-adoption.test.ts +118 -0
- package/src/store/__tests__/schema.test.ts +119 -0
- package/src/store/memory/disk.ts +17 -16
- package/src/store/run/disk.ts +169 -22
- package/src/store/schema.ts +135 -0
- package/src/store/session/disk.ts +30 -21
- package/src/store/task/__tests__/lookup-across-runs.test.ts +77 -0
- package/src/store/task/disk.ts +59 -15
- package/src/store/thread/disk.ts +21 -10
- package/src/streaming/coalesce.ts +1 -1
- package/src/telemetry/__tests__/latency-instruments.test.ts +120 -0
- package/src/telemetry/__tests__/metrics.test.ts +160 -0
- package/src/telemetry/__tests__/span-hierarchy.test.ts +94 -0
- package/src/telemetry/attributes.ts +109 -0
- package/src/telemetry/metrics.ts +190 -0
- package/src/test-setup.ts +3 -3
- package/src/test-support/paths.ts +17 -0
- package/src/tools/__tests__/path-containment.test.ts +167 -0
- package/src/tools/__tests__/posix-path.test.ts +83 -0
- package/src/tools/__tests__/sandboxed-search.test.ts +140 -0
- package/src/tools/builtins/__tests__/computer-use.test.ts +15 -2
- package/src/tools/builtins/__tests__/edit.test.ts +10 -270
- package/src/tools/builtins/__tests__/payload-budget.test.ts +20 -40
- package/src/tools/builtins/__tests__/read-file.test.ts +29 -1
- package/src/tools/builtins/__tests__/write-file.test.ts +13 -52
- package/src/tools/builtins/bash.ts +34 -4
- package/src/tools/builtins/computer-use.ts +7 -1
- package/src/tools/builtins/edit.ts +162 -116
- package/src/tools/builtins/glob.ts +71 -6
- package/src/tools/builtins/grep.ts +82 -8
- package/src/tools/builtins/index.ts +6 -6
- package/src/tools/builtins/ls.ts +4 -2
- package/src/tools/builtins/read-file.ts +40 -4
- package/src/tools/builtins/write-file.ts +44 -68
- package/src/tools/coordinator/__tests__/agent.test.ts +2 -2
- package/src/tools/coordinator/agent.ts +23 -10
- package/src/tools/coordinator/index.ts +111 -31
- package/src/tools/defineTool.ts +20 -0
- package/src/tools/glob-match.ts +50 -0
- package/src/tools/paths.ts +44 -0
- package/src/tools/posix-path.ts +73 -0
- package/src/types/advisory/config.ts +21 -3
- package/src/types/advisory/result.ts +12 -7
- package/src/types/agent/base.ts +22 -2
- package/src/types/agent/factory.ts +15 -9
- package/src/types/agent/gateway.ts +20 -0
- package/src/types/agent/manager.ts +19 -0
- package/src/types/agent/reactive.ts +50 -0
- package/src/types/agent/router.ts +11 -0
- package/src/types/agent/supervisor.ts +2 -2
- package/src/types/agent/task.ts +26 -0
- package/src/types/bidi/index.ts +119 -0
- package/src/types/common/__tests__/numeric-correctness.test.ts +123 -0
- package/src/types/common/index.ts +16 -4
- package/src/types/computer-use/index.ts +1 -1
- package/src/types/connector/mcp.ts +29 -0
- package/src/types/errors/__tests__/catalog.test.ts +130 -0
- package/src/types/errors/__tests__/errors.test.ts +118 -0
- package/src/types/errors/catalog.ts +211 -0
- package/src/types/errors/index.ts +121 -0
- package/src/types/guardrail/index.ts +62 -0
- package/src/types/hitl/index.ts +102 -20
- package/src/types/ids/index.ts +2 -2
- package/src/types/message/content.ts +50 -0
- package/src/types/message/index.ts +210 -9
- package/src/types/plugin/index.ts +82 -2
- package/src/types/provider/__tests__/cause-chain.test.ts +132 -0
- package/src/types/provider/__tests__/declared-retryable.test.ts +85 -0
- package/src/types/provider/__tests__/overflow-classification.test.ts +113 -0
- package/src/types/provider/chat.ts +29 -1
- package/src/types/provider/config.ts +83 -2
- package/src/types/provider/errors.ts +472 -0
- package/src/types/provider/index.ts +5 -0
- package/src/types/provider/stream.ts +63 -3
- package/src/types/rag/embedding.ts +11 -2
- package/src/types/run/__tests__/step.test.ts +123 -0
- package/src/types/run/answer-review.ts +51 -0
- package/src/types/run/config.ts +35 -0
- package/src/types/run/derive-status.ts +62 -0
- package/src/types/run/entity.ts +18 -0
- package/src/types/run/events.ts +230 -3
- package/src/types/run/index.ts +5 -0
- package/src/types/run/memory-promotion.ts +112 -0
- package/src/types/run/prepare-step.ts +109 -0
- package/src/types/run/state.ts +111 -0
- package/src/types/run/step.ts +95 -0
- package/src/types/run/stop-reason.ts +29 -1
- package/src/types/sandbox/index.ts +93 -12
- package/src/types/structured-output/index.ts +6 -5
- package/src/types/thread/entity.ts +2 -3
- package/src/types/tool/index.ts +204 -1
- package/src/types/tool/repair.ts +54 -0
- package/src/types/toolset/index.ts +2 -6
- package/src/utils/__tests__/atomic-write.test.ts +108 -0
- package/src/utils/__tests__/shell-compress.test.ts +49 -0
- package/src/utils/atomic-write.ts +85 -0
- package/src/utils/shell-compress.ts +12 -1
- package/src/verification/presets.ts +6 -3
|
@@ -0,0 +1,63 @@
|
|
|
1
|
+
import type { LLMProvider } from '../types/provider/interface.js';
|
|
2
|
+
import type { Scorer } from './types.js';
|
|
3
|
+
/**
|
|
4
|
+
* Grade an open-ended answer with a model.
|
|
5
|
+
*
|
|
6
|
+
* Every other scorer here is a pure function over the run, which is what
|
|
7
|
+
* makes them cheap and reproducible — and also what makes them unable to
|
|
8
|
+
* say anything about whether an answer is *good*. `containsScorer` can
|
|
9
|
+
* check that a required phrase appears; it cannot tell a correct
|
|
10
|
+
* explanation from a fluent wrong one. That was the hole: the dimension
|
|
11
|
+
* most worth guarding was the one with no scorer behind it.
|
|
12
|
+
*
|
|
13
|
+
* A judge is a network call, so this scorer can fail in ways a pure
|
|
14
|
+
* function cannot. It throws rather than returning a low score, and the
|
|
15
|
+
* harness files a throw as `unavailable` — a rate limit must not read as
|
|
16
|
+
* a regression. See `Score.unavailable`.
|
|
17
|
+
*/
|
|
18
|
+
export interface JudgeScorerConfig {
|
|
19
|
+
/** Defaults to `'judge'`. Give it a distinct name to run several. */
|
|
20
|
+
name?: string;
|
|
21
|
+
provider: LLMProvider;
|
|
22
|
+
model: string;
|
|
23
|
+
/**
|
|
24
|
+
* What "good" means, in the caller's words. REQUIRED.
|
|
25
|
+
*
|
|
26
|
+
* A judge asked to rate quality with no rubric rates fluency, which
|
|
27
|
+
* correlates with almost nothing worth measuring and drifts whenever
|
|
28
|
+
* the judge model changes. Making this optional would make the scorer
|
|
29
|
+
* easy to use and its output meaningless, so it is not optional.
|
|
30
|
+
*/
|
|
31
|
+
rubric: string;
|
|
32
|
+
/**
|
|
33
|
+
* Highest grade on the scale. Default 4.
|
|
34
|
+
*
|
|
35
|
+
* An integer scale, not a 0..1 float: models place a continuous score
|
|
36
|
+
* poorly and cluster on round numbers, while a short ordinal scale
|
|
37
|
+
* against a written rubric is a judgement they can actually make. The
|
|
38
|
+
* result is divided down to 0..1 for the report.
|
|
39
|
+
*
|
|
40
|
+
* The default is EVEN on purpose — an odd scale has a midpoint, and a
|
|
41
|
+
* midpoint is where an uncertain judge parks. Forcing a side produces a
|
|
42
|
+
* signal; a pile of 3-out-of-5s does not.
|
|
43
|
+
*/
|
|
44
|
+
scale?: number;
|
|
45
|
+
/**
|
|
46
|
+
* Show the judge which tools the run called. Default false.
|
|
47
|
+
*
|
|
48
|
+
* Useful when the rubric is about method rather than answer, and a
|
|
49
|
+
* needless cost otherwise — the trajectory is usually longer than the
|
|
50
|
+
* answer it produced.
|
|
51
|
+
*/
|
|
52
|
+
includeTrajectory?: boolean;
|
|
53
|
+
/**
|
|
54
|
+
* Cap on the answer text handed to the judge. Default 20000.
|
|
55
|
+
*
|
|
56
|
+
* Truncation is disclosed IN the prompt. A judge shown a silently cut
|
|
57
|
+
* answer marks it down for stopping mid-sentence, which scores our
|
|
58
|
+
* truncation rather than the run.
|
|
59
|
+
*/
|
|
60
|
+
maxOutputChars?: number;
|
|
61
|
+
}
|
|
62
|
+
export declare function judgeScorer(config: JudgeScorerConfig): Scorer;
|
|
63
|
+
//# sourceMappingURL=judge.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"judge.d.ts","sourceRoot":"","sources":["../../src/eval/judge.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,gCAAgC,CAAA;AACjE,OAAO,KAAK,EAA4B,MAAM,EAAE,MAAM,YAAY,CAAA;AAElE;;;;;;;;;;;;;;GAcG;AACH,MAAM,WAAW,iBAAiB;IACjC,qEAAqE;IACrE,IAAI,CAAC,EAAE,MAAM,CAAA;IACb,QAAQ,EAAE,WAAW,CAAA;IACrB,KAAK,EAAE,MAAM,CAAA;IACb;;;;;;;OAOG;IACH,MAAM,EAAE,MAAM,CAAA;IACd;;;;;;;;;;;OAWG;IACH,KAAK,CAAC,EAAE,MAAM,CAAA;IACd;;;;;;OAMG;IACH,iBAAiB,CAAC,EAAE,OAAO,CAAA;IAC3B;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,MAAM,CAAA;CACvB;AAwID,wBAAgB,WAAW,CAAC,MAAM,EAAE,iBAAiB,GAAG,MAAM,CAyC7D"}
|
|
@@ -0,0 +1,136 @@
|
|
|
1
|
+
import { collect } from '../provider/collect.js';
|
|
2
|
+
const DEFAULT_SCALE = 4;
|
|
3
|
+
const DEFAULT_MAX_OUTPUT_CHARS = 20_000;
|
|
4
|
+
/**
|
|
5
|
+
* Pull the verdict out of the reply.
|
|
6
|
+
*
|
|
7
|
+
* Models wrap JSON in prose or fences however firmly asked not to, so the
|
|
8
|
+
* first balanced object is extracted rather than the whole reply parsed.
|
|
9
|
+
* What is NOT tolerated is a missing or out-of-range grade: that is an
|
|
10
|
+
* unusable measurement, and the caller has to hear about it rather than
|
|
11
|
+
* receive a number invented from a fallback.
|
|
12
|
+
*/
|
|
13
|
+
function parseVerdict(reply, scale) {
|
|
14
|
+
const start = reply.indexOf('{');
|
|
15
|
+
if (start === -1)
|
|
16
|
+
throw new Error(`judge returned no JSON object: ${reply.slice(0, 200)}`);
|
|
17
|
+
let depth = 0;
|
|
18
|
+
let inString = false;
|
|
19
|
+
let escaped = false;
|
|
20
|
+
let end = -1;
|
|
21
|
+
for (let i = start; i < reply.length; i++) {
|
|
22
|
+
const ch = reply[i];
|
|
23
|
+
if (escaped) {
|
|
24
|
+
escaped = false;
|
|
25
|
+
continue;
|
|
26
|
+
}
|
|
27
|
+
if (ch === '\\') {
|
|
28
|
+
escaped = true;
|
|
29
|
+
continue;
|
|
30
|
+
}
|
|
31
|
+
if (ch === '"') {
|
|
32
|
+
inString = !inString;
|
|
33
|
+
continue;
|
|
34
|
+
}
|
|
35
|
+
if (inString)
|
|
36
|
+
continue;
|
|
37
|
+
if (ch === '{')
|
|
38
|
+
depth++;
|
|
39
|
+
else if (ch === '}') {
|
|
40
|
+
depth--;
|
|
41
|
+
if (depth === 0) {
|
|
42
|
+
end = i;
|
|
43
|
+
break;
|
|
44
|
+
}
|
|
45
|
+
}
|
|
46
|
+
}
|
|
47
|
+
if (end === -1)
|
|
48
|
+
throw new Error(`judge returned an unterminated JSON object: ${reply.slice(0, 200)}`);
|
|
49
|
+
let parsed;
|
|
50
|
+
try {
|
|
51
|
+
parsed = JSON.parse(reply.slice(start, end + 1));
|
|
52
|
+
}
|
|
53
|
+
catch (err) {
|
|
54
|
+
throw new Error(`judge returned unparsable JSON: ${err instanceof Error ? err.message : String(err)}`);
|
|
55
|
+
}
|
|
56
|
+
const grade = parsed.grade;
|
|
57
|
+
if (typeof grade !== 'number' || !Number.isFinite(grade)) {
|
|
58
|
+
throw new Error(`judge returned no numeric grade: ${JSON.stringify(parsed).slice(0, 200)}`);
|
|
59
|
+
}
|
|
60
|
+
if (grade < 0 || grade > scale) {
|
|
61
|
+
// Clamping instead would quietly turn a judge that misread the scale
|
|
62
|
+
// into a confident score, and a judge that misread the scale did not
|
|
63
|
+
// apply the rubric either.
|
|
64
|
+
throw new Error(`judge returned ${grade}, outside the 0..${scale} scale it was given`);
|
|
65
|
+
}
|
|
66
|
+
const reason = parsed.reason;
|
|
67
|
+
return {
|
|
68
|
+
grade,
|
|
69
|
+
reason: typeof reason === 'string' && reason.length > 0 ? reason : '(judge gave no reason)',
|
|
70
|
+
};
|
|
71
|
+
}
|
|
72
|
+
function buildPrompt(config, run, evalCase, scale) {
|
|
73
|
+
const limit = config.maxOutputChars ?? DEFAULT_MAX_OUTPUT_CHARS;
|
|
74
|
+
const answer = run.output ?? '';
|
|
75
|
+
const truncated = answer.length > limit;
|
|
76
|
+
const sections = [
|
|
77
|
+
'Grade the ANSWER against the RUBRIC.',
|
|
78
|
+
'',
|
|
79
|
+
`RUBRIC:\n${config.rubric}`,
|
|
80
|
+
'',
|
|
81
|
+
`TASK:\n${typeof evalCase.input === 'string' ? evalCase.input : JSON.stringify(evalCase.input)}`,
|
|
82
|
+
];
|
|
83
|
+
if (evalCase.expected !== undefined) {
|
|
84
|
+
sections.push('', `REFERENCE (one good answer, not the only one):\n${typeof evalCase.expected === 'string'
|
|
85
|
+
? evalCase.expected
|
|
86
|
+
: JSON.stringify(evalCase.expected)}`);
|
|
87
|
+
}
|
|
88
|
+
if (config.includeTrajectory === true) {
|
|
89
|
+
sections.push('', `TOOLS CALLED, in order:\n${run.toolCalls.length > 0 ? run.toolCalls.join(' -> ') : '(none)'}`);
|
|
90
|
+
}
|
|
91
|
+
sections.push('', `ANSWER:\n${truncated ? answer.slice(0, limit) : answer}`);
|
|
92
|
+
if (truncated) {
|
|
93
|
+
// Disclosed, so the judge does not mark the answer down for an
|
|
94
|
+
// ending we removed.
|
|
95
|
+
sections.push('', `(The answer above was cut at ${limit} characters by the harness, not by the agent. Do not penalise it for ending abruptly.)`);
|
|
96
|
+
}
|
|
97
|
+
sections.push('', `Reply with JSON and nothing else: {"grade": <integer 0..${scale}>, "reason": "<one or two sentences>"}.`, `0 means the rubric is not met at all; ${scale} means it is fully met. Judge only against the rubric.`);
|
|
98
|
+
return sections.join('\n');
|
|
99
|
+
}
|
|
100
|
+
export function judgeScorer(config) {
|
|
101
|
+
const scale = config.scale ?? DEFAULT_SCALE;
|
|
102
|
+
if (!Number.isInteger(scale) || scale < 1) {
|
|
103
|
+
throw new Error(`judgeScorer: scale must be a positive integer, got ${scale}`);
|
|
104
|
+
}
|
|
105
|
+
if (config.rubric.trim().length === 0) {
|
|
106
|
+
throw new Error('judgeScorer: a rubric is required. A judge with no rubric rates fluency, which is not what anyone means to measure.');
|
|
107
|
+
}
|
|
108
|
+
return {
|
|
109
|
+
name: config.name ?? 'judge',
|
|
110
|
+
async score(run, evalCase) {
|
|
111
|
+
const response = await collect(config.provider.chatStream({
|
|
112
|
+
model: config.model,
|
|
113
|
+
messages: [{ role: 'user', content: buildPrompt(config, run, evalCase, scale) }],
|
|
114
|
+
// The same run must grade the same way twice, or a
|
|
115
|
+
// regression cannot be told from sampling noise.
|
|
116
|
+
temperature: 0,
|
|
117
|
+
maxTokens: 512,
|
|
118
|
+
}));
|
|
119
|
+
const verdict = parseVerdict(response.message.content ?? '', scale);
|
|
120
|
+
return {
|
|
121
|
+
score: verdict.grade / scale,
|
|
122
|
+
reason: verdict.reason,
|
|
123
|
+
details: {
|
|
124
|
+
grade: verdict.grade,
|
|
125
|
+
scale,
|
|
126
|
+
// Carried so a suite can account for what the judging
|
|
127
|
+
// itself cost. A judge is the most expensive scorer there
|
|
128
|
+
// is, and a bill nobody can attribute is a bill nobody
|
|
129
|
+
// controls.
|
|
130
|
+
judgeTokens: response.usage.totalTokens,
|
|
131
|
+
},
|
|
132
|
+
};
|
|
133
|
+
},
|
|
134
|
+
};
|
|
135
|
+
}
|
|
136
|
+
//# sourceMappingURL=judge.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"judge.js","sourceRoot":"","sources":["../../src/eval/judge.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,MAAM,wBAAwB,CAAA;AAgEhD,MAAM,aAAa,GAAG,CAAC,CAAA;AACvB,MAAM,wBAAwB,GAAG,MAAM,CAAA;AAOvC;;;;;;;;GAQG;AACH,SAAS,YAAY,CAAC,KAAa,EAAE,KAAa;IACjD,MAAM,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;IAChC,IAAI,KAAK,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,kCAAkC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAE1F,IAAI,KAAK,GAAG,CAAC,CAAA;IACb,IAAI,QAAQ,GAAG,KAAK,CAAA;IACpB,IAAI,OAAO,GAAG,KAAK,CAAA;IACnB,IAAI,GAAG,GAAG,CAAC,CAAC,CAAA;IACZ,KAAK,IAAI,CAAC,GAAG,KAAK,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3C,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;QACnB,IAAI,OAAO,EAAE,CAAC;YACb,OAAO,GAAG,KAAK,CAAA;YACf,SAAQ;QACT,CAAC;QACD,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;YACjB,OAAO,GAAG,IAAI,CAAA;YACd,SAAQ;QACT,CAAC;QACD,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YAChB,QAAQ,GAAG,CAAC,QAAQ,CAAA;YACpB,SAAQ;QACT,CAAC;QACD,IAAI,QAAQ;YAAE,SAAQ;QACtB,IAAI,EAAE,KAAK,GAAG;YAAE,KAAK,EAAE,CAAA;aAClB,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YACrB,KAAK,EAAE,CAAA;YACP,IAAI,KAAK,KAAK,CAAC,EAAE,CAAC;gBACjB,GAAG,GAAG,CAAC,CAAA;gBACP,MAAK;YACN,CAAC;QACF,CAAC;IACF,CAAC;IACD,IAAI,GAAG,KAAK,CAAC,CAAC;QACb,MAAM,IAAI,KAAK,CAAC,+CAA+C,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAEtF,IAAI,MAAe,CAAA;IACnB,IAAI,CAAC;QACJ,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC,CAAA;IACjD,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACd,MAAM,IAAI,KAAK,CACd,mCAAmC,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CACrF,CAAA;IACF,CAAC;IAED,MAAM,KAAK,GAAI,MAA8B,CAAC,KAAK,CAAA;IACnD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;QAC1D,MAAM,IAAI,KAAK,CAAC,oCAAoC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAC5F,CAAC;IACD,IAAI,KAAK,GAAG,CAAC,IAAI,KAAK,GAAG,KAAK,EAAE,CAAC;QAChC,qEAAqE;QACrE,qEAAqE;QACrE,2BAA2B;QAC3B,MAAM,IAAI,KAAK,CAAC,kBAAkB,KAAK,oBAAoB,KAAK,qBAAqB,CAAC,CAAA;IACvF,CAAC;IAED,MAAM,MAAM,GAAI,MAA+B,CAAC,MAAM,CAAA;IACtD,OAAO;QACN,KAAK;QACL,MAAM,EAAE,OAAO,MAAM,KAAK,QAAQ,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,wBAAwB;KAC3F,CAAA;AACF,CAAC;AAED,SAAS,WAAW,CACnB,MAAyB,EACzB,GAAY,EACZ,QAAkB,EAClB,KAAa;IAEb,MAAM,KAAK,GAAG,MAAM,CAAC,cAAc,IAAI,wBAAwB,CAAA;IAC/D,MAAM,MAAM,GAAG,GAAG,CAAC,MAAM,IAAI,EAAE,CAAA;IAC/B,MAAM,SAAS,GAAG,MAAM,CAAC,MAAM,GAAG,KAAK,CAAA;IAEvC,MAAM,QAAQ,GAAa;QAC1B,sCAAsC;QACtC,EAAE;QACF,YAAY,MAAM,CAAC,MAAM,EAAE;QAC3B,EAAE;QACF,UAAU,OAAO,QAAQ,CAAC,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE;KAChG,CAAA;IAED,IAAI,QAAQ,CAAC,QAAQ,KAAK,SAAS,EAAE,CAAC;QACrC,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,mDACC,OAAO,QAAQ,CAAC,QAAQ,KAAK,QAAQ;YACpC,CAAC,CAAC,QAAQ,CAAC,QAAQ;YACnB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,QAAQ,CACpC,EAAE,CACF,CAAA;IACF,CAAC;IAED,IAAI,MAAM,CAAC,iBAAiB,KAAK,IAAI,EAAE,CAAC;QACvC,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,4BAA4B,GAAG,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,QAAQ,EAAE,CAC9F,CAAA;IACF,CAAC;IAED,QAAQ,CAAC,IAAI,CAAC,EAAE,EAAE,YAAY,SAAS,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAA;IAC5E,IAAI,SAAS,EAAE,CAAC;QACf,+DAA+D;QAC/D,qBAAqB;QACrB,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,gCAAgC,KAAK,wFAAwF,CAC7H,CAAA;IACF,CAAC;IAED,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,2DAA2D,KAAK,yCAAyC,EACzG,yCAAyC,KAAK,wDAAwD,CACtG,CAAA;IAED,OAAO,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;AAC3B,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,MAAyB;IACpD,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,IAAI,aAAa,CAAA;IAC3C,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,KAAK,CAAC,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;QAC3C,MAAM,IAAI,KAAK,CAAC,sDAAsD,KAAK,EAAE,CAAC,CAAA;IAC/E,CAAC;IACD,IAAI,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACvC,MAAM,IAAI,KAAK,CACd,qHAAqH,CACrH,CAAA;IACF,CAAC;IAED,OAAO;QACN,IAAI,EAAE,MAAM,CAAC,IAAI,IAAI,OAAO;QAC5B,KAAK,CAAC,KAAK,CAAC,GAAY,EAAE,QAAkB;YAC3C,MAAM,QAAQ,GAAG,MAAM,OAAO,CAC7B,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC;gBAC1B,KAAK,EAAE,MAAM,CAAC,KAAK;gBACnB,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,WAAW,CAAC,MAAM,EAAE,GAAG,EAAE,QAAQ,EAAE,KAAK,CAAC,EAAE,CAAC;gBAChF,mDAAmD;gBACnD,iDAAiD;gBACjD,WAAW,EAAE,CAAC;gBACd,SAAS,EAAE,GAAG;aACd,CAAC,CACF,CAAA;YAED,MAAM,OAAO,GAAG,YAAY,CAAC,QAAQ,CAAC,OAAO,CAAC,OAAO,IAAI,EAAE,EAAE,KAAK,CAAC,CAAA;YACnE,OAAO;gBACN,KAAK,EAAE,OAAO,CAAC,KAAK,GAAG,KAAK;gBAC5B,MAAM,EAAE,OAAO,CAAC,MAAM;gBACtB,OAAO,EAAE;oBACR,KAAK,EAAE,OAAO,CAAC,KAAK;oBACpB,KAAK;oBACL,sDAAsD;oBACtD,0DAA0D;oBAC1D,uDAAuD;oBACvD,YAAY;oBACZ,WAAW,EAAE,QAAQ,CAAC,KAAK,CAAC,WAAW;iBACvC;aACD,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC"}
|
|
@@ -0,0 +1,40 @@
|
|
|
1
|
+
import type { EvalCase, EvalRun, Score, Scorer } from './types.js';
|
|
2
|
+
/**
|
|
3
|
+
* How closely the run's tool sequence matched the expected one, as F1 over
|
|
4
|
+
* the longest common subsequence.
|
|
5
|
+
*
|
|
6
|
+
* Trajectory, not final answer. Namzu's most load-bearing behavior is
|
|
7
|
+
* tuned by constants nobody could measure — `search_tools` activates the
|
|
8
|
+
* top 5 deferred tools, compaction fires at 0.7, six state lists cap at
|
|
9
|
+
* 25 — and changing any of them, or a tool description, or the
|
|
10
|
+
* deferred-tools prompt block, could silently make the agent take four
|
|
11
|
+
* tool calls where it took one. Final-answer scoring cannot see that;
|
|
12
|
+
* this can.
|
|
13
|
+
*
|
|
14
|
+
* Extra calls cut precision, missing calls cut recall, so a run that does
|
|
15
|
+
* the right thing wastefully and a run that skips a step score
|
|
16
|
+
* differently — which is the distinction a final-answer score collapses.
|
|
17
|
+
*/
|
|
18
|
+
export declare function trajectoryScorer(): Scorer;
|
|
19
|
+
/** The run settled cleanly rather than erroring or being cut off. */
|
|
20
|
+
export declare function completionScorer(acceptable?: readonly string[]): Scorer;
|
|
21
|
+
/**
|
|
22
|
+
* The run stayed within a step budget.
|
|
23
|
+
*
|
|
24
|
+
* A regression that makes the agent take four turns where it took one is
|
|
25
|
+
* invisible to correctness scoring and very visible on the bill.
|
|
26
|
+
*/
|
|
27
|
+
export declare function stepBudgetScorer(maxSteps: number): Scorer;
|
|
28
|
+
/** The final text contains every required substring. */
|
|
29
|
+
export declare function containsScorer(...required: string[]): Scorer;
|
|
30
|
+
/**
|
|
31
|
+
* Judge the run with a caller-supplied predicate.
|
|
32
|
+
*
|
|
33
|
+
* The escape hatch for anything the built-in scorers do not cover. For
|
|
34
|
+
* a model-graded judge, reach for `judgeScorer` first: it handles the
|
|
35
|
+
* parts that are easy to get wrong — a rubric it refuses to run without,
|
|
36
|
+
* an ordinal scale rather than a float, disclosed truncation, and a
|
|
37
|
+
* failed call that reports as unjudged rather than as a zero.
|
|
38
|
+
*/
|
|
39
|
+
export declare function customScorer(name: string, fn: (run: EvalRun, evalCase: EvalCase) => Score | Promise<Score>): Scorer;
|
|
40
|
+
//# sourceMappingURL=scorers.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"scorers.d.ts","sourceRoot":"","sources":["../../src/eval/scorers.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,YAAY,CAAA;AAyBlE;;;;;;;;;;;;;;;GAeG;AACH,wBAAgB,gBAAgB,IAAI,MAAM,CAwCzC;AAED,qEAAqE;AACrE,wBAAgB,gBAAgB,CAC/B,UAAU,GAAE,SAAS,MAAM,EAAmC,GAC5D,MAAM,CAqBR;AAED;;;;;GAKG;AACH,wBAAgB,gBAAgB,CAAC,QAAQ,EAAE,MAAM,GAAG,MAAM,CAezD;AAED,wDAAwD;AACxD,wBAAgB,cAAc,CAAC,GAAG,QAAQ,EAAE,MAAM,EAAE,GAAG,MAAM,CAmB5D;AAED;;;;;;;;GAQG;AACH,wBAAgB,YAAY,CAC3B,IAAI,EAAE,MAAM,EACZ,EAAE,EAAE,CAAC,GAAG,EAAE,OAAO,EAAE,QAAQ,EAAE,QAAQ,KAAK,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,GAC9D,MAAM,CAER"}
|
|
@@ -0,0 +1,151 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Longest common subsequence length between two tool sequences.
|
|
3
|
+
*
|
|
4
|
+
* Subsequence, not set intersection: order carries meaning in a
|
|
5
|
+
* trajectory. Reading a file before editing it is not the same run as
|
|
6
|
+
* editing then reading, and a set-based score cannot tell them apart.
|
|
7
|
+
*/
|
|
8
|
+
function lcsLength(a, b) {
|
|
9
|
+
const table = Array.from({ length: a.length + 1 }, () => new Array(b.length + 1).fill(0));
|
|
10
|
+
for (let i = 1; i <= a.length; i++) {
|
|
11
|
+
for (let j = 1; j <= b.length; j++) {
|
|
12
|
+
const row = table[i];
|
|
13
|
+
const prev = table[i - 1];
|
|
14
|
+
if (!row || !prev)
|
|
15
|
+
continue;
|
|
16
|
+
row[j] =
|
|
17
|
+
a[i - 1] === b[j - 1] ? (prev[j - 1] ?? 0) + 1 : Math.max(prev[j] ?? 0, row[j - 1] ?? 0);
|
|
18
|
+
}
|
|
19
|
+
}
|
|
20
|
+
return table[a.length]?.[b.length] ?? 0;
|
|
21
|
+
}
|
|
22
|
+
/**
|
|
23
|
+
* How closely the run's tool sequence matched the expected one, as F1 over
|
|
24
|
+
* the longest common subsequence.
|
|
25
|
+
*
|
|
26
|
+
* Trajectory, not final answer. Namzu's most load-bearing behavior is
|
|
27
|
+
* tuned by constants nobody could measure — `search_tools` activates the
|
|
28
|
+
* top 5 deferred tools, compaction fires at 0.7, six state lists cap at
|
|
29
|
+
* 25 — and changing any of them, or a tool description, or the
|
|
30
|
+
* deferred-tools prompt block, could silently make the agent take four
|
|
31
|
+
* tool calls where it took one. Final-answer scoring cannot see that;
|
|
32
|
+
* this can.
|
|
33
|
+
*
|
|
34
|
+
* Extra calls cut precision, missing calls cut recall, so a run that does
|
|
35
|
+
* the right thing wastefully and a run that skips a step score
|
|
36
|
+
* differently — which is the distinction a final-answer score collapses.
|
|
37
|
+
*/
|
|
38
|
+
export function trajectoryScorer() {
|
|
39
|
+
return {
|
|
40
|
+
name: 'trajectory',
|
|
41
|
+
score(run, evalCase) {
|
|
42
|
+
const expected = evalCase.expectedTools ?? [];
|
|
43
|
+
const actual = run.toolCalls;
|
|
44
|
+
if (expected.length === 0 && actual.length === 0) {
|
|
45
|
+
return { score: 1, reason: 'no tools expected and none called' };
|
|
46
|
+
}
|
|
47
|
+
if (expected.length === 0) {
|
|
48
|
+
return {
|
|
49
|
+
score: 0,
|
|
50
|
+
reason: `expected no tool calls, got ${actual.length}: ${actual.join(' → ')}`,
|
|
51
|
+
details: { actual },
|
|
52
|
+
};
|
|
53
|
+
}
|
|
54
|
+
if (actual.length === 0) {
|
|
55
|
+
return {
|
|
56
|
+
score: 0,
|
|
57
|
+
reason: `expected ${expected.join(' → ')}, but no tools were called`,
|
|
58
|
+
details: { expected },
|
|
59
|
+
};
|
|
60
|
+
}
|
|
61
|
+
const matched = lcsLength(expected, actual);
|
|
62
|
+
const precision = matched / actual.length;
|
|
63
|
+
const recall = matched / expected.length;
|
|
64
|
+
const f1 = precision + recall === 0 ? 0 : (2 * precision * recall) / (precision + recall);
|
|
65
|
+
return {
|
|
66
|
+
score: f1,
|
|
67
|
+
reason: f1 === 1
|
|
68
|
+
? `exact trajectory: ${actual.join(' → ')}`
|
|
69
|
+
: `matched ${matched}/${expected.length} in order; expected ${expected.join(' → ')}, got ${actual.join(' → ')}`,
|
|
70
|
+
details: { precision, recall, matched, expected, actual },
|
|
71
|
+
};
|
|
72
|
+
},
|
|
73
|
+
};
|
|
74
|
+
}
|
|
75
|
+
/** The run settled cleanly rather than erroring or being cut off. */
|
|
76
|
+
export function completionScorer(acceptable = ['end_turn', 'stop_condition']) {
|
|
77
|
+
return {
|
|
78
|
+
name: 'completion',
|
|
79
|
+
// Binary by construction: the run either settled cleanly or it did
|
|
80
|
+
// not, so there is no score between 0 and 1 for a mean to soften.
|
|
81
|
+
// Averaged in, a hard 0 here is carried by three good fuzzy scores.
|
|
82
|
+
severity: 'gate',
|
|
83
|
+
score(run) {
|
|
84
|
+
if (run.error) {
|
|
85
|
+
return { score: 0, reason: `run failed: ${run.error}` };
|
|
86
|
+
}
|
|
87
|
+
const ok = run.stopReason !== undefined && acceptable.includes(run.stopReason);
|
|
88
|
+
return {
|
|
89
|
+
score: ok ? 1 : 0,
|
|
90
|
+
reason: ok
|
|
91
|
+
? `settled as ${run.stopReason}`
|
|
92
|
+
: `settled as ${run.stopReason ?? 'unknown'}, expected one of ${acceptable.join(', ')}`,
|
|
93
|
+
details: { stopReason: run.stopReason },
|
|
94
|
+
};
|
|
95
|
+
},
|
|
96
|
+
};
|
|
97
|
+
}
|
|
98
|
+
/**
|
|
99
|
+
* The run stayed within a step budget.
|
|
100
|
+
*
|
|
101
|
+
* A regression that makes the agent take four turns where it took one is
|
|
102
|
+
* invisible to correctness scoring and very visible on the bill.
|
|
103
|
+
*/
|
|
104
|
+
export function stepBudgetScorer(maxSteps) {
|
|
105
|
+
return {
|
|
106
|
+
name: 'step-budget',
|
|
107
|
+
score(run) {
|
|
108
|
+
const used = run.steps.length;
|
|
109
|
+
return {
|
|
110
|
+
score: used <= maxSteps ? 1 : Math.max(0, maxSteps / used),
|
|
111
|
+
reason: used <= maxSteps
|
|
112
|
+
? `${used}/${maxSteps} steps`
|
|
113
|
+
: `over budget: ${used} steps against a ${maxSteps} allowance`,
|
|
114
|
+
details: { used, maxSteps },
|
|
115
|
+
};
|
|
116
|
+
},
|
|
117
|
+
};
|
|
118
|
+
}
|
|
119
|
+
/** The final text contains every required substring. */
|
|
120
|
+
export function containsScorer(...required) {
|
|
121
|
+
return {
|
|
122
|
+
name: 'contains',
|
|
123
|
+
// A required fragment is required. Partial credit here reads as
|
|
124
|
+
// "most of the answer was there", which is not what was asked.
|
|
125
|
+
severity: 'gate',
|
|
126
|
+
score(run) {
|
|
127
|
+
const text = run.output ?? '';
|
|
128
|
+
const missing = required.filter((r) => !text.includes(r));
|
|
129
|
+
return {
|
|
130
|
+
score: required.length === 0 ? 1 : (required.length - missing.length) / required.length,
|
|
131
|
+
reason: missing.length === 0
|
|
132
|
+
? `found all ${required.length} required fragments`
|
|
133
|
+
: `missing: ${missing.join(', ')}`,
|
|
134
|
+
details: { missing },
|
|
135
|
+
};
|
|
136
|
+
},
|
|
137
|
+
};
|
|
138
|
+
}
|
|
139
|
+
/**
|
|
140
|
+
* Judge the run with a caller-supplied predicate.
|
|
141
|
+
*
|
|
142
|
+
* The escape hatch for anything the built-in scorers do not cover. For
|
|
143
|
+
* a model-graded judge, reach for `judgeScorer` first: it handles the
|
|
144
|
+
* parts that are easy to get wrong — a rubric it refuses to run without,
|
|
145
|
+
* an ordinal scale rather than a float, disclosed truncation, and a
|
|
146
|
+
* failed call that reports as unjudged rather than as a zero.
|
|
147
|
+
*/
|
|
148
|
+
export function customScorer(name, fn) {
|
|
149
|
+
return { name, score: fn };
|
|
150
|
+
}
|
|
151
|
+
//# sourceMappingURL=scorers.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"scorers.js","sourceRoot":"","sources":["../../src/eval/scorers.ts"],"names":[],"mappings":"AAEA;;;;;;GAMG;AACH,SAAS,SAAS,CAAC,CAAoB,EAAE,CAAoB;IAC5D,MAAM,KAAK,GAAe,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,MAAM,GAAG,CAAC,EAAE,EAAE,GAAG,EAAE,CACnE,IAAI,KAAK,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAC/B,CAAA;IACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACpC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACpC,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;YACpB,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAA;YACzB,IAAI,CAAC,GAAG,IAAI,CAAC,IAAI;gBAAE,SAAQ;YAC3B,GAAG,CAAC,CAAC,CAAC;gBACL,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAA;QAC1F,CAAC;IACF,CAAC;IACD,OAAO,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAA;AACxC,CAAC;AAED;;;;;;;;;;;;;;;GAeG;AACH,MAAM,UAAU,gBAAgB;IAC/B,OAAO;QACN,IAAI,EAAE,YAAY;QAClB,KAAK,CAAC,GAAY,EAAE,QAAkB;YACrC,MAAM,QAAQ,GAAG,QAAQ,CAAC,aAAa,IAAI,EAAE,CAAA;YAC7C,MAAM,MAAM,GAAG,GAAG,CAAC,SAAS,CAAA;YAE5B,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAClD,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,mCAAmC,EAAE,CAAA;YACjE,CAAC;YACD,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAC3B,OAAO;oBACN,KAAK,EAAE,CAAC;oBACR,MAAM,EAAE,+BAA+B,MAAM,CAAC,MAAM,KAAK,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;oBAC7E,OAAO,EAAE,EAAE,MAAM,EAAE;iBACnB,CAAA;YACF,CAAC;YACD,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBACzB,OAAO;oBACN,KAAK,EAAE,CAAC;oBACR,MAAM,EAAE,YAAY,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,4BAA4B;oBACpE,OAAO,EAAE,EAAE,QAAQ,EAAE;iBACrB,CAAA;YACF,CAAC;YAED,MAAM,OAAO,GAAG,SAAS,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAA;YAC3C,MAAM,SAAS,GAAG,OAAO,GAAG,MAAM,CAAC,MAAM,CAAA;YACzC,MAAM,MAAM,GAAG,OAAO,GAAG,QAAQ,CAAC,MAAM,CAAA;YACxC,MAAM,EAAE,GAAG,SAAS,GAAG,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,SAAS,GAAG,MAAM,CAAC,GAAG,CAAC,SAAS,GAAG,MAAM,CAAC,CAAA;YAEzF,OAAO;gBACN,KAAK,EAAE,EAAE;gBACT,MAAM,EACL,EAAE,KAAK,CAAC;oBACP,CAAC,CAAC,qBAAqB,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;oBAC3C,CAAC,CAAC,WAAW,OAAO,IAAI,QAAQ,CAAC,MAAM,uBAAuB,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,SAAS,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;gBACjH,OAAO,EAAE,EAAE,SAAS,EAAE,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE;aACzD,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED,qEAAqE;AACrE,MAAM,UAAU,gBAAgB,CAC/B,aAAgC,CAAC,UAAU,EAAE,gBAAgB,CAAC;IAE9D,OAAO;QACN,IAAI,EAAE,YAAY;QAClB,mEAAmE;QACnE,kEAAkE;QAClE,oEAAoE;QACpE,QAAQ,EAAE,MAAM;QAChB,KAAK,CAAC,GAAY;YACjB,IAAI,GAAG,CAAC,KAAK,EAAE,CAAC;gBACf,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,eAAe,GAAG,CAAC,KAAK,EAAE,EAAE,CAAA;YACxD,CAAC;YACD,MAAM,EAAE,GAAG,GAAG,CAAC,UAAU,KAAK,SAAS,IAAI,UAAU,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC,CAAA;YAC9E,OAAO;gBACN,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBACjB,MAAM,EAAE,EAAE;oBACT,CAAC,CAAC,cAAc,GAAG,CAAC,UAAU,EAAE;oBAChC,CAAC,CAAC,cAAc,GAAG,CAAC,UAAU,IAAI,SAAS,qBAAqB,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;gBACxF,OAAO,EAAE,EAAE,UAAU,EAAE,GAAG,CAAC,UAAU,EAAE;aACvC,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,gBAAgB,CAAC,QAAgB;IAChD,OAAO;QACN,IAAI,EAAE,aAAa;QACnB,KAAK,CAAC,GAAY;YACjB,MAAM,IAAI,GAAG,GAAG,CAAC,KAAK,CAAC,MAAM,CAAA;YAC7B,OAAO;gBACN,KAAK,EAAE,IAAI,IAAI,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,QAAQ,GAAG,IAAI,CAAC;gBAC1D,MAAM,EACL,IAAI,IAAI,QAAQ;oBACf,CAAC,CAAC,GAAG,IAAI,IAAI,QAAQ,QAAQ;oBAC7B,CAAC,CAAC,gBAAgB,IAAI,oBAAoB,QAAQ,YAAY;gBAChE,OAAO,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE;aAC3B,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED,wDAAwD;AACxD,MAAM,UAAU,cAAc,CAAC,GAAG,QAAkB;IACnD,OAAO;QACN,IAAI,EAAE,UAAU;QAChB,gEAAgE;QAChE,+DAA+D;QAC/D,QAAQ,EAAE,MAAM;QAChB,KAAK,CAAC,GAAY;YACjB,MAAM,IAAI,GAAG,GAAG,CAAC,MAAM,IAAI,EAAE,CAAA;YAC7B,MAAM,OAAO,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAA;YACzD,OAAO;gBACN,KAAK,EAAE,QAAQ,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,GAAG,QAAQ,CAAC,MAAM;gBACvF,MAAM,EACL,OAAO,CAAC,MAAM,KAAK,CAAC;oBACnB,CAAC,CAAC,aAAa,QAAQ,CAAC,MAAM,qBAAqB;oBACnD,CAAC,CAAC,YAAY,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;gBACpC,OAAO,EAAE,EAAE,OAAO,EAAE;aACpB,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;;;;;;;GAQG;AACH,MAAM,UAAU,YAAY,CAC3B,IAAY,EACZ,EAAgE;IAEhE,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,EAAE,EAAE,CAAA;AAC3B,CAAC"}
|
|
@@ -0,0 +1,150 @@
|
|
|
1
|
+
import type { StepResult } from '../types/run/step.js';
|
|
2
|
+
/**
|
|
3
|
+
* One case in a dataset: an input, and what a good run looks like.
|
|
4
|
+
*
|
|
5
|
+
* `TInput` is the agent's own input type, so a dataset is checked against
|
|
6
|
+
* the thing it tests rather than against `unknown`.
|
|
7
|
+
*/
|
|
8
|
+
export interface EvalCase<TInput = unknown> {
|
|
9
|
+
name: string;
|
|
10
|
+
input: TInput;
|
|
11
|
+
/** Free-form expectation, consumed by whichever scorers you attach. */
|
|
12
|
+
expected?: unknown;
|
|
13
|
+
/**
|
|
14
|
+
* Tool names the run is expected to call, in order. Consumed by
|
|
15
|
+
* `trajectoryScorer`.
|
|
16
|
+
*/
|
|
17
|
+
expectedTools?: readonly string[];
|
|
18
|
+
/** Per-case scorer overrides; falls back to the dataset's scorers. */
|
|
19
|
+
scorers?: readonly Scorer[];
|
|
20
|
+
}
|
|
21
|
+
/** What actually happened, handed to every scorer. */
|
|
22
|
+
export interface EvalRun {
|
|
23
|
+
/** Final assistant text, when the run produced one. */
|
|
24
|
+
output: string | null;
|
|
25
|
+
/** Schema-validated final output, when `structuredOutput` was requested. */
|
|
26
|
+
structuredOutput?: unknown;
|
|
27
|
+
steps: readonly StepResult[];
|
|
28
|
+
/** Tool names in call order, flattened across steps. */
|
|
29
|
+
toolCalls: readonly string[];
|
|
30
|
+
stopReason?: string;
|
|
31
|
+
totalTokens: number;
|
|
32
|
+
totalCostUsd: number;
|
|
33
|
+
durationMs: number;
|
|
34
|
+
/** Set when the run threw rather than settling. */
|
|
35
|
+
error?: string;
|
|
36
|
+
}
|
|
37
|
+
/**
|
|
38
|
+
* A judgement about one run.
|
|
39
|
+
*
|
|
40
|
+
* `reason` is REQUIRED, not optional. A bare number tells you a run got
|
|
41
|
+
* worse without telling you how, which is exactly when you need to know —
|
|
42
|
+
* so every scorer has to be able to explain itself, and one that cannot
|
|
43
|
+
* is a scorer that should not exist.
|
|
44
|
+
*/
|
|
45
|
+
export interface Score {
|
|
46
|
+
/** 0..1. Higher is better. Ignored when `unavailable` is set. */
|
|
47
|
+
score: number;
|
|
48
|
+
reason: string;
|
|
49
|
+
/** Anything the scorer wants to surface (counts, diffs, matched items). */
|
|
50
|
+
details?: Record<string, unknown>;
|
|
51
|
+
/**
|
|
52
|
+
* The scorer could not produce a judgement — it threw, its provider was
|
|
53
|
+
* unreachable, its verdict did not parse.
|
|
54
|
+
*
|
|
55
|
+
* A failed measurement is not a measurement of zero. Scoring it zero
|
|
56
|
+
* says "the run was bad" when the truth is "we do not know", and the
|
|
57
|
+
* two demand opposite responses: one is a regression to chase, the
|
|
58
|
+
* other is a broken harness to fix. This mattered little while every
|
|
59
|
+
* scorer was a pure function — a throw there is a bug — and matters a
|
|
60
|
+
* great deal now that a scorer can be a network call, where one 429
|
|
61
|
+
* would otherwise turn a green suite red and send somebody hunting a
|
|
62
|
+
* regression that never happened.
|
|
63
|
+
*
|
|
64
|
+
* An unavailable score is excluded from the case mean's numerator AND
|
|
65
|
+
* its denominator, and the case it belongs to is reported as
|
|
66
|
+
* inconclusive rather than passed or failed.
|
|
67
|
+
*/
|
|
68
|
+
unavailable?: boolean;
|
|
69
|
+
}
|
|
70
|
+
/**
|
|
71
|
+
* How a scorer's verdict affects the case's.
|
|
72
|
+
*
|
|
73
|
+
* A single unweighted mean over every scorer, compared to one suite-wide
|
|
74
|
+
* threshold, made the two halves of the design fight each other. At the
|
|
75
|
+
* default threshold of 1 the harness never reports a false pass — but a
|
|
76
|
+
* trajectory F1 and a graded judge can essentially never reach 1, so every
|
|
77
|
+
* real suite lowers it. And every step down buys tolerance for the fuzzy
|
|
78
|
+
* scorers by buying the same tolerance for the deterministic ones: at 0.75,
|
|
79
|
+
* a trajectory score of 0 alongside three perfect scores averages to 0.75
|
|
80
|
+
* and reports **passed**. The exact regression the harness exists to catch
|
|
81
|
+
* comes back green.
|
|
82
|
+
*
|
|
83
|
+
* - `gate` — a miss fails the case outright, whatever the mean says.
|
|
84
|
+
* For the deterministic checks: it called the wrong tools, it did not
|
|
85
|
+
* finish, the required phrase is absent.
|
|
86
|
+
* - `soft` — contributes to the mean only. For the fuzzy ones, where a
|
|
87
|
+
* number below 1 is normal and a threshold is a judgement call.
|
|
88
|
+
*
|
|
89
|
+
* Defaults to `soft`, so a suite that sets nothing behaves exactly as
|
|
90
|
+
* before.
|
|
91
|
+
*/
|
|
92
|
+
export type ScorerSeverity = 'gate' | 'soft';
|
|
93
|
+
export interface Scorer {
|
|
94
|
+
name: string;
|
|
95
|
+
score(run: EvalRun, evalCase: EvalCase): Score | Promise<Score>;
|
|
96
|
+
/** See {@link ScorerSeverity}. Default `soft`. */
|
|
97
|
+
severity?: ScorerSeverity;
|
|
98
|
+
/**
|
|
99
|
+
* The score this scorer must reach. Defaults to the experiment's
|
|
100
|
+
* `passThreshold`.
|
|
101
|
+
*
|
|
102
|
+
* Per-scorer because "good enough" is not one number across dimensions:
|
|
103
|
+
* a trajectory match at 0.8 may be fine while a completion check at 0.8
|
|
104
|
+
* is meaningless — it either finished or it did not.
|
|
105
|
+
*/
|
|
106
|
+
threshold?: number;
|
|
107
|
+
}
|
|
108
|
+
/**
|
|
109
|
+
* `inconclusive` is a first-class outcome: every scorer that ran was
|
|
110
|
+
* unavailable, so there is no evidence either way. Folding it into
|
|
111
|
+
* `failed` invents a regression and folding it into `passed` invents
|
|
112
|
+
* evidence; both are worse than saying nothing was measured.
|
|
113
|
+
*/
|
|
114
|
+
export type CaseStatus = 'passed' | 'failed' | 'inconclusive';
|
|
115
|
+
export interface CaseResult {
|
|
116
|
+
case: string;
|
|
117
|
+
run: EvalRun;
|
|
118
|
+
scores: Record<string, Score>;
|
|
119
|
+
/** Mean of this case's AVAILABLE scores. Zero when none were. */
|
|
120
|
+
mean: number;
|
|
121
|
+
status: CaseStatus;
|
|
122
|
+
/** True only for `status: 'passed'`. */
|
|
123
|
+
passed: boolean;
|
|
124
|
+
/**
|
|
125
|
+
* Gate scorers that missed their threshold, by name.
|
|
126
|
+
*
|
|
127
|
+
* Named rather than counted: a case reported failed with a mean of 0.75
|
|
128
|
+
* otherwise sends somebody to read four scores and guess which one
|
|
129
|
+
* mattered. Absent when no gate missed.
|
|
130
|
+
*/
|
|
131
|
+
failedGates?: readonly string[];
|
|
132
|
+
}
|
|
133
|
+
export interface ExperimentReport {
|
|
134
|
+
name: string;
|
|
135
|
+
cases: readonly CaseResult[];
|
|
136
|
+
/** Mean score across the cases that produced one. */
|
|
137
|
+
mean: number;
|
|
138
|
+
passed: number;
|
|
139
|
+
failed: number;
|
|
140
|
+
/** Cases where no scorer could produce a judgement. */
|
|
141
|
+
inconclusive: number;
|
|
142
|
+
/**
|
|
143
|
+
* Per-scorer means over the cases each one actually judged, for
|
|
144
|
+
* spotting which dimension regressed. A scorer that was unavailable on
|
|
145
|
+
* some cases is averaged over the rest rather than dragged toward zero.
|
|
146
|
+
*/
|
|
147
|
+
byScorer: Record<string, number>;
|
|
148
|
+
durationMs: number;
|
|
149
|
+
}
|
|
150
|
+
//# sourceMappingURL=types.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../src/eval/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,sBAAsB,CAAA;AAEtD;;;;;GAKG;AACH,MAAM,WAAW,QAAQ,CAAC,MAAM,GAAG,OAAO;IACzC,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,EAAE,MAAM,CAAA;IACb,uEAAuE;IACvE,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB;;;OAGG;IACH,aAAa,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;IACjC,sEAAsE;IACtE,OAAO,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;CAC3B;AAED,sDAAsD;AACtD,MAAM,WAAW,OAAO;IACvB,uDAAuD;IACvD,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,4EAA4E;IAC5E,gBAAgB,CAAC,EAAE,OAAO,CAAA;IAC1B,KAAK,EAAE,SAAS,UAAU,EAAE,CAAA;IAC5B,wDAAwD;IACxD,SAAS,EAAE,SAAS,MAAM,EAAE,CAAA;IAC5B,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,WAAW,EAAE,MAAM,CAAA;IACnB,YAAY,EAAE,MAAM,CAAA;IACpB,UAAU,EAAE,MAAM,CAAA;IAClB,mDAAmD;IACnD,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED;;;;;;;GAOG;AACH,MAAM,WAAW,KAAK;IACrB,iEAAiE;IACjE,KAAK,EAAE,MAAM,CAAA;IACb,MAAM,EAAE,MAAM,CAAA;IACd,2EAA2E;IAC3E,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAA;IACjC;;;;;;;;;;;;;;;;OAgBG;IACH,WAAW,CAAC,EAAE,OAAO,CAAA;CACrB;AAED;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,MAAM,MAAM,cAAc,GAAG,MAAM,GAAG,MAAM,CAAA;AAE5C,MAAM,WAAW,MAAM;IACtB,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,CAAC,GAAG,EAAE,OAAO,EAAE,QAAQ,EAAE,QAAQ,GAAG,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,CAAA;IAC/D,kDAAkD;IAClD,QAAQ,CAAC,EAAE,cAAc,CAAA;IACzB;;;;;;;OAOG;IACH,SAAS,CAAC,EAAE,MAAM,CAAA;CAClB;AAED;;;;;GAKG;AACH,MAAM,MAAM,UAAU,GAAG,QAAQ,GAAG,QAAQ,GAAG,cAAc,CAAA;AAE7D,MAAM,WAAW,UAAU;IAC1B,IAAI,EAAE,MAAM,CAAA;IACZ,GAAG,EAAE,OAAO,CAAA;IACZ,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,KAAK,CAAC,CAAA;IAC7B,iEAAiE;IACjE,IAAI,EAAE,MAAM,CAAA;IACZ,MAAM,EAAE,UAAU,CAAA;IAClB,wCAAwC;IACxC,MAAM,EAAE,OAAO,CAAA;IACf;;;;;;OAMG;IACH,WAAW,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;CAC/B;AAED,MAAM,WAAW,gBAAgB;IAChC,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,EAAE,SAAS,UAAU,EAAE,CAAA;IAC5B,qDAAqD;IACrD,IAAI,EAAE,MAAM,CAAA;IACZ,MAAM,EAAE,MAAM,CAAA;IACd,MAAM,EAAE,MAAM,CAAA;IACd,uDAAuD;IACvD,YAAY,EAAE,MAAM,CAAA;IACpB;;;;OAIG;IACH,QAAQ,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,UAAU,EAAE,MAAM,CAAA;CAClB"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/eval/types.ts"],"names":[],"mappings":""}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"local.d.ts","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,6BAA6B,CAAA;AAC9E,OAAO,KAAK,EACX,eAAe,EACf,cAAc,EACd,aAAa,EACb,mBAAmB,EACnB,oBAAoB,EACpB,MAAM,6BAA6B,CAAA;AACpC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAEhD,MAAM,WAAW,4BAA4B;IAC5C,EAAE,EAAE,MAAM,CAAA;IACV,GAAG,EAAE,MAAM,CAAA;IACX,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,YAAY,CAAC,EAAE,mBAAmB,EAAE,CAAA;IACpC,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED,qBAAa,qBAAsB,SAAQ,oBAAqB,YAAW,eAAe;IACzF,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAA;IACnB,QAAQ,CAAC,WAAW,EAAE,oBAAoB,CAAU;IAEpD,OAAO,CAAC,GAAG,CAAQ;IACnB,OAAO,CAAC,QAAQ,CAAS;IACzB,OAAO,CAAC,OAAO,CAAwB;IACvC,OAAO,CAAC,YAAY,CAAuB;IAC3C,OAAO,CAAC,KAAK,CAAoB;gBAErB,OAAO,EAAE,4BAA4B;cAUjC,YAAY,IAAI,OAAO,CAAC,IAAI,CAAC;cAO7B,UAAU,IAAI,OAAO,CAAC,IAAI,CAAC;IAE3C,MAAM,IAAI,MAAM;IAIhB,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,IAAI;IAQ5B,WAAW,IAAI,OAAO;IAItB,WAAW,CAAC,YAAY,EAAE,MAAM,GAAG,MAAM;IAIzC,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAI1C,UAAU,IAAI,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAIpC,eAAe,IAAI,mBAAmB,EAAE;IAIxC,aAAa,CAAC,GAAG,EAAE,mBAAmB,GAAG,OAAO;IAI1C,cAAc,CACnB,OAAO,EAAE,MAAM,EACf,IAAI,GAAE,MAAM,EAAO,EACnB,OAAO,CAAC,EAAE,cAAc,GACtB,OAAO,CAAC,aAAa,CAAC;
|
|
1
|
+
{"version":3,"file":"local.d.ts","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,6BAA6B,CAAA;AAC9E,OAAO,KAAK,EACX,eAAe,EACf,cAAc,EACd,aAAa,EACb,mBAAmB,EACnB,oBAAoB,EACpB,MAAM,6BAA6B,CAAA;AACpC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAEhD,MAAM,WAAW,4BAA4B;IAC5C,EAAE,EAAE,MAAM,CAAA;IACV,GAAG,EAAE,MAAM,CAAA;IACX,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,YAAY,CAAC,EAAE,mBAAmB,EAAE,CAAA;IACpC,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED,qBAAa,qBAAsB,SAAQ,oBAAqB,YAAW,eAAe;IACzF,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAA;IACnB,QAAQ,CAAC,WAAW,EAAE,oBAAoB,CAAU;IAEpD,OAAO,CAAC,GAAG,CAAQ;IACnB,OAAO,CAAC,QAAQ,CAAS;IACzB,OAAO,CAAC,OAAO,CAAwB;IACvC,OAAO,CAAC,YAAY,CAAuB;IAC3C,OAAO,CAAC,KAAK,CAAoB;gBAErB,OAAO,EAAE,4BAA4B;cAUjC,YAAY,IAAI,OAAO,CAAC,IAAI,CAAC;cAO7B,UAAU,IAAI,OAAO,CAAC,IAAI,CAAC;IAE3C,MAAM,IAAI,MAAM;IAIhB,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,IAAI;IAQ5B,WAAW,IAAI,OAAO;IAItB,WAAW,CAAC,YAAY,EAAE,MAAM,GAAG,MAAM;IAIzC,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAI1C,UAAU,IAAI,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAIpC,eAAe,IAAI,mBAAmB,EAAE;IAIxC,aAAa,CAAC,GAAG,EAAE,mBAAmB,GAAG,OAAO;IAI1C,cAAc,CACnB,OAAO,EAAE,MAAM,EACf,IAAI,GAAE,MAAM,EAAO,EACnB,OAAO,CAAC,EAAE,cAAc,GACtB,OAAO,CAAC,aAAa,CAAC;IA4DzB,QAAQ,IAAI,2BAA2B;CAWvC"}
|
package/dist/execution/local.js
CHANGED
|
@@ -66,7 +66,12 @@ export class LocalExecutionContext extends BaseExecutionContext {
|
|
|
66
66
|
const cwd = options?.cwd ? resolve(this.cwd, options.cwd) : this.cwd;
|
|
67
67
|
const env = { ...process.env, ...this.envVars, ...options?.env };
|
|
68
68
|
const timeoutMs = options?.timeoutMs ?? 30_000;
|
|
69
|
-
|
|
69
|
+
// node's spawn() only keeps argv separation when it execs the binary directly;
|
|
70
|
+
// with shell enabled it re-joins command+args into one unescaped string and hands
|
|
71
|
+
// it to sh/cmd.exe, so any caller-supplied arg becomes shell syntax rather than a
|
|
72
|
+
// literal value. Since `command`/`args` here are whatever a tool call passed in,
|
|
73
|
+
// shell interpretation must be an explicit opt-in, never the silent default.
|
|
74
|
+
const shell = options?.shell ?? this.shell ?? false;
|
|
70
75
|
return new Promise((resolvePromise) => {
|
|
71
76
|
const start = performance.now();
|
|
72
77
|
const proc = spawn(command, args, {
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"local.js","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAA;AAC1C,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAA;AACpC,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAA;AASnC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAWhD,MAAM,OAAO,qBAAsB,SAAQ,oBAAoB;IACrD,EAAE,CAAQ;IACV,WAAW,GAAyB,OAAO,CAAA;IAE5C,GAAG,CAAQ;IACX,QAAQ,CAAS;IACjB,OAAO,CAAwB;IAC/B,YAAY,CAAuB;IACnC,KAAK,CAAoB;IAEjC,YAAY,OAAqC;QAChD,KAAK,EAAE,CAAA;QACP,IAAI,CAAC,EAAE,GAAG,OAAO,CAAC,EAAE,CAAA;QACpB,IAAI,CAAC,GAAG,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;QAC/B,IAAI,CAAC,QAAQ,GAAG,OAAO,CAAC,QAAQ,IAAI,IAAI,CAAA;QACxC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC,OAAO,IAAI,EAAE,CAAA;QACpC,IAAI,CAAC,YAAY,GAAG,OAAO,CAAC,YAAY,IAAI,CAAC,YAAY,EAAE,SAAS,EAAE,OAAO,CAAC,CAAA;QAC9E,IAAI,CAAC,KAAK,GAAG,OAAO,CAAC,KAAK,CAAA;IAC3B,CAAC;IAES,KAAK,CAAC,YAAY;QAC3B,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,gCAAgC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;IAC1D,CAAC;IAES,KAAK,CAAC,UAAU,KAAmB,CAAC;IAE9C,MAAM;QACL,OAAO,IAAI,CAAC,GAAG,CAAA;IAChB,CAAC;IAED,MAAM,CAAC,MAAc;QACpB,MAAM,QAAQ,GAAG,OAAO,CAAC,MAAM,CAAC,CAAA;QAChC,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,QAAQ,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,GAAG,QAAQ,CAAA;IACpB,CAAC;IAED,WAAW;QACV,OAAO,IAAI,CAAC,QAAQ,CAAA;IACrB,CAAC;IAED,WAAW,CAAC,YAAoB;QAC/B,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,YAAY,CAAC,CAAA;IACvC,CAAC;IAED,SAAS,CAAC,GAAW;QACpB,OAAO,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;IAC7C,CAAC;IAED,UAAU;QACT,OAAO,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAA;IAC3B,CAAC;IAED,eAAe;QACd,OAAO,CAAC,GAAG,IAAI,CAAC,YAAY,CAAC,CAAA;IAC9B,CAAC;IAED,aAAa,CAAC,GAAwB;QACrC,OAAO,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAA;IACvC,CAAC;IAED,KAAK,CAAC,cAAc,CACnB,OAAe,EACf,OAAiB,EAAE,EACnB,OAAwB;QAExB,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,OAAO,CAAC,EAAE,CAAC;YACpE,OAAO;gBACN,QAAQ,EAAE,CAAC;gBACX,MAAM,EAAE,EAAE;gBACV,MAAM,EAAE,yEAAyE;gBACjF,UAAU,EAAE,CAAC;aACb,CAAA;QACF,CAAC;QAED,MAAM,GAAG,GAAG,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAA;QACpE,MAAM,GAAG,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,GAAG,OAAO,EAAE,GAAG,EAAE,CAAA;QAChE,MAAM,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,MAAM,CAAA;QAC9C,MAAM,KAAK,GAAG,OAAO,EAAE,KAAK,IAAI,IAAI,CAAC,KAAK,IAAI,
|
|
1
|
+
{"version":3,"file":"local.js","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAA;AAC1C,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAA;AACpC,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAA;AASnC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAWhD,MAAM,OAAO,qBAAsB,SAAQ,oBAAoB;IACrD,EAAE,CAAQ;IACV,WAAW,GAAyB,OAAO,CAAA;IAE5C,GAAG,CAAQ;IACX,QAAQ,CAAS;IACjB,OAAO,CAAwB;IAC/B,YAAY,CAAuB;IACnC,KAAK,CAAoB;IAEjC,YAAY,OAAqC;QAChD,KAAK,EAAE,CAAA;QACP,IAAI,CAAC,EAAE,GAAG,OAAO,CAAC,EAAE,CAAA;QACpB,IAAI,CAAC,GAAG,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;QAC/B,IAAI,CAAC,QAAQ,GAAG,OAAO,CAAC,QAAQ,IAAI,IAAI,CAAA;QACxC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC,OAAO,IAAI,EAAE,CAAA;QACpC,IAAI,CAAC,YAAY,GAAG,OAAO,CAAC,YAAY,IAAI,CAAC,YAAY,EAAE,SAAS,EAAE,OAAO,CAAC,CAAA;QAC9E,IAAI,CAAC,KAAK,GAAG,OAAO,CAAC,KAAK,CAAA;IAC3B,CAAC;IAES,KAAK,CAAC,YAAY;QAC3B,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,gCAAgC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;IAC1D,CAAC;IAES,KAAK,CAAC,UAAU,KAAmB,CAAC;IAE9C,MAAM;QACL,OAAO,IAAI,CAAC,GAAG,CAAA;IAChB,CAAC;IAED,MAAM,CAAC,MAAc;QACpB,MAAM,QAAQ,GAAG,OAAO,CAAC,MAAM,CAAC,CAAA;QAChC,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,QAAQ,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,GAAG,QAAQ,CAAA;IACpB,CAAC;IAED,WAAW;QACV,OAAO,IAAI,CAAC,QAAQ,CAAA;IACrB,CAAC;IAED,WAAW,CAAC,YAAoB;QAC/B,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,YAAY,CAAC,CAAA;IACvC,CAAC;IAED,SAAS,CAAC,GAAW;QACpB,OAAO,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;IAC7C,CAAC;IAED,UAAU;QACT,OAAO,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAA;IAC3B,CAAC;IAED,eAAe;QACd,OAAO,CAAC,GAAG,IAAI,CAAC,YAAY,CAAC,CAAA;IAC9B,CAAC;IAED,aAAa,CAAC,GAAwB;QACrC,OAAO,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAA;IACvC,CAAC;IAED,KAAK,CAAC,cAAc,CACnB,OAAe,EACf,OAAiB,EAAE,EACnB,OAAwB;QAExB,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,OAAO,CAAC,EAAE,CAAC;YACpE,OAAO;gBACN,QAAQ,EAAE,CAAC;gBACX,MAAM,EAAE,EAAE;gBACV,MAAM,EAAE,yEAAyE;gBACjF,UAAU,EAAE,CAAC;aACb,CAAA;QACF,CAAC;QAED,MAAM,GAAG,GAAG,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAA;QACpE,MAAM,GAAG,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,GAAG,OAAO,EAAE,GAAG,EAAE,CAAA;QAChE,MAAM,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,MAAM,CAAA;QAC9C,+EAA+E;QAC/E,kFAAkF;QAClF,kFAAkF;QAClF,iFAAiF;QACjF,6EAA6E;QAC7E,MAAM,KAAK,GAAG,OAAO,EAAE,KAAK,IAAI,IAAI,CAAC,KAAK,IAAI,KAAK,CAAA;QAEnD,OAAO,IAAI,OAAO,CAAgB,CAAC,cAAc,EAAE,EAAE;YACpD,MAAM,KAAK,GAAG,WAAW,CAAC,GAAG,EAAE,CAAA;YAC/B,MAAM,IAAI,GAAG,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE;gBACjC,GAAG;gBACH,GAAG;gBACH,KAAK;gBACL,OAAO,EAAE,SAAS;aAClB,CAAC,CAAA;YAEF,IAAI,MAAM,GAAG,EAAE,CAAA;YACf,IAAI,MAAM,GAAG,EAAE,CAAA;YAEf,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE;gBACzB,cAAc,CAAC;oBACd,QAAQ,EAAE,IAAI,IAAI,CAAC;oBACnB,MAAM;oBACN,MAAM;oBACN,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE;gBACxB,cAAc,CAAC;oBACd,QAAQ,EAAE,CAAC;oBACX,MAAM;oBACN,MAAM,EAAE,GAAG,CAAC,OAAO;oBACnB,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;QACH,CAAC,CAAC,CAAA;IACH,CAAC;IAED,QAAQ;QACP,OAAO;YACN,EAAE,EAAE,IAAI,CAAC,EAAE;YACX,WAAW,EAAE,OAAO;YACpB,GAAG,EAAE,IAAI,CAAC,GAAG;YACb,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,YAAY,EAAE,IAAI,CAAC,YAAY;YAC/B,KAAK,EAAE,IAAI,CAAC,KAAK;SACjB,CAAA;IACF,CAAC;CACD"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"local.test.d.ts","sourceRoot":"","sources":["../../src/execution/local.test.ts"],"names":[],"mappings":""}
|