@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 +1,85 @@
|
|
|
1
|
+
import { describe, expect, it } from 'vitest';
|
|
2
|
+
import { runExperiment } from '../experiment.js';
|
|
3
|
+
import { containsScorer, stepBudgetScorer, trajectoryScorer } from '../scorers.js';
|
|
4
|
+
/**
|
|
5
|
+
* An eval harness that reports green on a broken suite is worse than no
|
|
6
|
+
* harness: it converts "we do not know" into "we checked". Both defects
|
|
7
|
+
* here did exactly that.
|
|
8
|
+
*/
|
|
9
|
+
const emptyRun = () => ({
|
|
10
|
+
output: 'ok',
|
|
11
|
+
steps: [],
|
|
12
|
+
toolCalls: [],
|
|
13
|
+
totalTokens: 0,
|
|
14
|
+
totalCostUsd: 0,
|
|
15
|
+
durationMs: 0,
|
|
16
|
+
});
|
|
17
|
+
describe('a case whose run THREW cannot score above zero', () => {
|
|
18
|
+
it('scores 0 instead of a perfect step budget', async () => {
|
|
19
|
+
// `executeCase` catches the throw and returns an empty run, and an
|
|
20
|
+
// empty run walks into every scorer's happy path: 0 steps is within
|
|
21
|
+
// any allowance, and "no tools expected, none called" is a match. The
|
|
22
|
+
// suite reported 1.0 for a case that never ran.
|
|
23
|
+
const report = await runExperiment({
|
|
24
|
+
name: 'crash',
|
|
25
|
+
cases: [{ name: 'boom', input: 'x' }],
|
|
26
|
+
scorers: [stepBudgetScorer(5), trajectoryScorer()],
|
|
27
|
+
run: () => {
|
|
28
|
+
throw new Error('provider exploded');
|
|
29
|
+
},
|
|
30
|
+
});
|
|
31
|
+
expect(report.cases[0]?.scores['step-budget']?.score).toBe(0);
|
|
32
|
+
expect(report.cases[0]?.scores.trajectory?.score).toBe(0);
|
|
33
|
+
expect(report.mean).toBe(0);
|
|
34
|
+
});
|
|
35
|
+
it('says why, so the green-to-red change is diagnosable', async () => {
|
|
36
|
+
const report = await runExperiment({
|
|
37
|
+
name: 'crash',
|
|
38
|
+
cases: [{ name: 'boom', input: 'x' }],
|
|
39
|
+
scorers: [stepBudgetScorer(5)],
|
|
40
|
+
run: () => {
|
|
41
|
+
throw new Error('provider exploded');
|
|
42
|
+
},
|
|
43
|
+
});
|
|
44
|
+
expect(report.cases[0]?.scores['step-budget']?.reason).toContain('provider exploded');
|
|
45
|
+
});
|
|
46
|
+
it('still scores a healthy run normally', async () => {
|
|
47
|
+
const report = await runExperiment({
|
|
48
|
+
name: 'fine',
|
|
49
|
+
cases: [{ name: 'ok', input: 'x' }],
|
|
50
|
+
scorers: [stepBudgetScorer(5)],
|
|
51
|
+
run: () => Promise.resolve(emptyRun()),
|
|
52
|
+
});
|
|
53
|
+
expect(report.cases[0]?.scores['step-budget']?.score).toBe(1);
|
|
54
|
+
});
|
|
55
|
+
});
|
|
56
|
+
describe('two scorers cannot silently collapse into one', () => {
|
|
57
|
+
it('refuses a duplicate scorer name rather than overwriting', async () => {
|
|
58
|
+
// Scores are keyed by name, so two `containsScorer(...)` instances —
|
|
59
|
+
// both called 'contains' — meant the second overwrote the first and
|
|
60
|
+
// the case mean was computed over the wrong denominator. With 'a'
|
|
61
|
+
// scoring 0 and 'b' scoring 1 the suite reported 1.0 where the honest
|
|
62
|
+
// answer is 0.5.
|
|
63
|
+
await expect(runExperiment({
|
|
64
|
+
name: 'dupe',
|
|
65
|
+
cases: [{ name: 'c', input: 'x' }],
|
|
66
|
+
scorers: [containsScorer('a'), containsScorer('b')],
|
|
67
|
+
run: () => Promise.resolve({ ...emptyRun(), output: 'b' }),
|
|
68
|
+
})).rejects.toThrow(/Duplicate scorer name "contains"/);
|
|
69
|
+
});
|
|
70
|
+
it('accepts distinct names', async () => {
|
|
71
|
+
const report = await runExperiment({
|
|
72
|
+
name: 'ok',
|
|
73
|
+
cases: [{ name: 'c', input: 'x' }],
|
|
74
|
+
scorers: [
|
|
75
|
+
{ ...containsScorer('a'), name: 'contains-a' },
|
|
76
|
+
{ ...containsScorer('b'), name: 'contains-b' },
|
|
77
|
+
],
|
|
78
|
+
run: () => Promise.resolve({ ...emptyRun(), output: 'b' }),
|
|
79
|
+
});
|
|
80
|
+
expect(report.cases[0]?.scores['contains-a']?.score).toBe(0);
|
|
81
|
+
expect(report.cases[0]?.scores['contains-b']?.score).toBe(1);
|
|
82
|
+
expect(report.cases[0]?.mean).toBe(0.5);
|
|
83
|
+
});
|
|
84
|
+
});
|
|
85
|
+
//# sourceMappingURL=harness-honesty.test.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"harness-honesty.test.js","sourceRoot":"","sources":["../../../src/eval/__tests__/harness-honesty.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAE7C,OAAO,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAA;AAChD,OAAO,EAAE,cAAc,EAAE,gBAAgB,EAAE,gBAAgB,EAAE,MAAM,eAAe,CAAA;AAGlF;;;;GAIG;AAEH,MAAM,QAAQ,GAAG,GAAY,EAAE,CAAC,CAAC;IAChC,MAAM,EAAE,IAAI;IACZ,KAAK,EAAE,EAAE;IACT,SAAS,EAAE,EAAE;IACb,WAAW,EAAE,CAAC;IACd,YAAY,EAAE,CAAC;IACf,UAAU,EAAE,CAAC;CACb,CAAC,CAAA;AAEF,QAAQ,CAAC,gDAAgD,EAAE,GAAG,EAAE;IAC/D,EAAE,CAAC,2CAA2C,EAAE,KAAK,IAAI,EAAE;QAC1D,mEAAmE;QACnE,oEAAoE;QACpE,sEAAsE;QACtE,gDAAgD;QAChD,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,OAAO;YACb,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC;YACrC,OAAO,EAAE,CAAC,gBAAgB,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,CAAC;YAClD,GAAG,EAAE,GAAG,EAAE;gBACT,MAAM,IAAI,KAAK,CAAC,mBAAmB,CAAC,CAAA;YACrC,CAAC;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,aAAa,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7D,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,UAAU,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACzD,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC5B,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,qDAAqD,EAAE,KAAK,IAAI,EAAE;QACpE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,OAAO;YACb,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC;YACrC,OAAO,EAAE,CAAC,gBAAgB,CAAC,CAAC,CAAC,CAAC;YAC9B,GAAG,EAAE,GAAG,EAAE;gBACT,MAAM,IAAI,KAAK,CAAC,mBAAmB,CAAC,CAAA;YACrC,CAAC;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,aAAa,CAAC,EAAE,MAAM,CAAC,CAAC,SAAS,CAAC,mBAAmB,CAAC,CAAA;IACtF,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,qCAAqC,EAAE,KAAK,IAAI,EAAE;QACpD,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC;YACnC,OAAO,EAAE,CAAC,gBAAgB,CAAC,CAAC,CAAC,CAAC;YAC9B,GAAG,EAAE,GAAG,EAAE,CAAC,OAAO,CAAC,OAAO,CAAC,QAAQ,EAAE,CAAC;SACtC,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,aAAa,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC9D,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,+CAA+C,EAAE,GAAG,EAAE;IAC9D,EAAE,CAAC,yDAAyD,EAAE,KAAK,IAAI,EAAE;QACxE,qEAAqE;QACrE,oEAAoE;QACpE,kEAAkE;QAClE,sEAAsE;QACtE,iBAAiB;QACjB,MAAM,MAAM,CACX,aAAa,CAAC;YACb,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC;YAClC,OAAO,EAAE,CAAC,cAAc,CAAC,GAAG,CAAC,EAAE,cAAc,CAAC,GAAG,CAAC,CAAC;YACnD,GAAG,EAAE,GAAG,EAAE,CAAC,OAAO,CAAC,OAAO,CAAC,EAAE,GAAG,QAAQ,EAAE,EAAE,MAAM,EAAE,GAAG,EAAE,CAAC;SAC1D,CAAC,CACF,CAAC,OAAO,CAAC,OAAO,CAAC,kCAAkC,CAAC,CAAA;IACtD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,wBAAwB,EAAE,KAAK,IAAI,EAAE;QACvC,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,IAAI;YACV,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC;YAClC,OAAO,EAAE;gBACR,EAAE,GAAG,cAAc,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE;gBAC9C,EAAE,GAAG,cAAc,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE;aAC9C;YACD,GAAG,EAAE,GAAG,EAAE,CAAC,OAAO,CAAC,OAAO,CAAC,EAAE,GAAG,QAAQ,EAAE,EAAE,MAAM,EAAE,GAAG,EAAE,CAAC;SAC1D,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,YAAY,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC5D,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,YAAY,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC5D,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;IACxC,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"harness.test.d.ts","sourceRoot":"","sources":["../../../src/eval/__tests__/harness.test.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,207 @@
|
|
|
1
|
+
import { describe, expect, it } from 'vitest';
|
|
2
|
+
import { formatReport, runExperiment } from '../experiment.js';
|
|
3
|
+
import { completionScorer, containsScorer, customScorer, stepBudgetScorer, trajectoryScorer, } from '../scorers.js';
|
|
4
|
+
/**
|
|
5
|
+
* There was no evaluation harness of any kind — no dataset, no scorer, no
|
|
6
|
+
* judge, no trajectory assertion. So the constants that govern namzu's
|
|
7
|
+
* most load-bearing behavior (`search_tools` top-k, the compaction
|
|
8
|
+
* threshold, six state-list caps, every builtin tool description) were
|
|
9
|
+
* tunable but not measurable: change one and the only way to learn the
|
|
10
|
+
* agent now takes four tool calls where it took one was for a user to
|
|
11
|
+
* hit it.
|
|
12
|
+
*/
|
|
13
|
+
function run(overrides = {}) {
|
|
14
|
+
return {
|
|
15
|
+
output: 'done',
|
|
16
|
+
steps: [],
|
|
17
|
+
toolCalls: [],
|
|
18
|
+
stopReason: 'end_turn',
|
|
19
|
+
totalTokens: 0,
|
|
20
|
+
totalCostUsd: 0,
|
|
21
|
+
durationMs: 0,
|
|
22
|
+
...overrides,
|
|
23
|
+
};
|
|
24
|
+
}
|
|
25
|
+
const CASE = { name: 'c', input: null };
|
|
26
|
+
describe('trajectoryScorer', () => {
|
|
27
|
+
// `Scorer.score` may be async (a model-graded judge is just an async
|
|
28
|
+
// predicate), so every call site awaits even when the built-in is sync.
|
|
29
|
+
const score = async (expectedTools, toolCalls) => await trajectoryScorer().score(run({ toolCalls }), { ...CASE, expectedTools });
|
|
30
|
+
it('scores an exact trajectory 1', async () => {
|
|
31
|
+
expect((await score(['read', 'edit'], ['read', 'edit'])).score).toBe(1);
|
|
32
|
+
});
|
|
33
|
+
it('distinguishes ORDER — a set-based score could not', async () => {
|
|
34
|
+
// Reading before editing is not the same run as editing then reading.
|
|
35
|
+
const reversed = await score(['read', 'edit'], ['edit', 'read']);
|
|
36
|
+
expect(reversed.score).toBeLessThan(1);
|
|
37
|
+
expect(reversed.score).toBeGreaterThan(0);
|
|
38
|
+
});
|
|
39
|
+
it('penalises an extra call through precision', async () => {
|
|
40
|
+
const wasteful = await score(['read'], ['read', 'read', 'read']);
|
|
41
|
+
expect(wasteful.score).toBeLessThan(1);
|
|
42
|
+
expect(wasteful.details?.recall).toBe(1);
|
|
43
|
+
expect(wasteful.details?.precision).toBeLessThan(1);
|
|
44
|
+
});
|
|
45
|
+
it('penalises a missing call through recall', async () => {
|
|
46
|
+
const lazy = await score(['read', 'edit', 'bash'], ['read']);
|
|
47
|
+
expect(lazy.details?.precision).toBe(1);
|
|
48
|
+
expect(lazy.details?.recall).toBeLessThan(1);
|
|
49
|
+
});
|
|
50
|
+
it('separates "wasteful" from "skipped" — a final-answer score collapses them', async () => {
|
|
51
|
+
const wasteful = await score(['read', 'edit'], ['read', 'ls', 'edit']);
|
|
52
|
+
const skipped = await score(['read', 'edit'], ['read']);
|
|
53
|
+
expect(wasteful.score).not.toBe(skipped.score);
|
|
54
|
+
});
|
|
55
|
+
it('scores 1 when no tools were expected and none ran', async () => {
|
|
56
|
+
expect((await score([], [])).score).toBe(1);
|
|
57
|
+
});
|
|
58
|
+
it('scores 0 when tools were expected and none ran', async () => {
|
|
59
|
+
const s = await score(['read'], []);
|
|
60
|
+
expect(s.score).toBe(0);
|
|
61
|
+
expect(s.reason).toContain('no tools were called');
|
|
62
|
+
});
|
|
63
|
+
it('scores 0 when none were expected and some ran', async () => {
|
|
64
|
+
expect((await score([], ['bash'])).score).toBe(0);
|
|
65
|
+
});
|
|
66
|
+
it('always explains itself', async () => {
|
|
67
|
+
expect((await score(['read'], ['edit'])).reason.length).toBeGreaterThan(0);
|
|
68
|
+
});
|
|
69
|
+
});
|
|
70
|
+
describe('completionScorer', () => {
|
|
71
|
+
it('accepts a clean settle', async () => {
|
|
72
|
+
expect((await completionScorer().score(run({ stopReason: 'end_turn' }), CASE)).score).toBe(1);
|
|
73
|
+
});
|
|
74
|
+
it('rejects a budget-forced stop', async () => {
|
|
75
|
+
const s = await completionScorer().score(run({ stopReason: 'max_iterations' }), CASE);
|
|
76
|
+
expect(s.score).toBe(0);
|
|
77
|
+
expect(s.reason).toContain('max_iterations');
|
|
78
|
+
});
|
|
79
|
+
it('reports the error when the run threw', async () => {
|
|
80
|
+
const s = await completionScorer().score(run({ error: 'boom' }), CASE);
|
|
81
|
+
expect(s.score).toBe(0);
|
|
82
|
+
expect(s.reason).toContain('boom');
|
|
83
|
+
});
|
|
84
|
+
});
|
|
85
|
+
describe('stepBudgetScorer', () => {
|
|
86
|
+
const withSteps = (n) => run({ steps: Array.from({ length: n }, () => ({})) });
|
|
87
|
+
it('passes within budget', async () => {
|
|
88
|
+
expect((await stepBudgetScorer(5).score(withSteps(3), CASE)).score).toBe(1);
|
|
89
|
+
});
|
|
90
|
+
it('degrades proportionally past it, rather than snapping to zero', async () => {
|
|
91
|
+
// A run that took 6 steps against a 5 allowance is worse than one
|
|
92
|
+
// that took 20; a binary score cannot say so.
|
|
93
|
+
const slight = (await stepBudgetScorer(5).score(withSteps(6), CASE)).score;
|
|
94
|
+
const severe = (await stepBudgetScorer(5).score(withSteps(20), CASE)).score;
|
|
95
|
+
expect(slight).toBeGreaterThan(severe);
|
|
96
|
+
expect(slight).toBeLessThan(1);
|
|
97
|
+
});
|
|
98
|
+
});
|
|
99
|
+
describe('containsScorer', () => {
|
|
100
|
+
it('scores the fraction found and names what is missing', async () => {
|
|
101
|
+
const s = await containsScorer('alpha', 'beta').score(run({ output: 'alpha only' }), CASE);
|
|
102
|
+
expect(s.score).toBe(0.5);
|
|
103
|
+
expect(s.reason).toContain('beta');
|
|
104
|
+
});
|
|
105
|
+
});
|
|
106
|
+
describe('runExperiment', () => {
|
|
107
|
+
const cases = [
|
|
108
|
+
{ name: 'good', input: 'a', expectedTools: ['read'] },
|
|
109
|
+
{ name: 'bad', input: 'b', expectedTools: ['read', 'edit'] },
|
|
110
|
+
];
|
|
111
|
+
it('scores every case and aggregates', async () => {
|
|
112
|
+
const report = await runExperiment({
|
|
113
|
+
name: 'demo',
|
|
114
|
+
cases,
|
|
115
|
+
scorers: [trajectoryScorer()],
|
|
116
|
+
run: async (input) => run({ toolCalls: input === 'a' ? ['read'] : ['bash'] }),
|
|
117
|
+
});
|
|
118
|
+
expect(report.cases).toHaveLength(2);
|
|
119
|
+
expect(report.passed).toBe(1);
|
|
120
|
+
expect(report.failed).toBe(1);
|
|
121
|
+
expect(report.byScorer.trajectory).toBeCloseTo(0.5, 5);
|
|
122
|
+
});
|
|
123
|
+
it('a throwing case is a RESULT, not a crash', async () => {
|
|
124
|
+
// A suite whose first broken case aborts tells you nothing about the
|
|
125
|
+
// other forty.
|
|
126
|
+
const report = await runExperiment({
|
|
127
|
+
name: 'demo',
|
|
128
|
+
cases,
|
|
129
|
+
scorers: [completionScorer()],
|
|
130
|
+
run: async (input) => {
|
|
131
|
+
if (input === 'a')
|
|
132
|
+
throw new Error('exploded');
|
|
133
|
+
return run();
|
|
134
|
+
},
|
|
135
|
+
});
|
|
136
|
+
expect(report.cases).toHaveLength(2);
|
|
137
|
+
expect(report.cases[0]?.run.error).toBe('exploded');
|
|
138
|
+
expect(report.cases[1]?.passed).toBe(true);
|
|
139
|
+
});
|
|
140
|
+
it('a throwing scorer scores zero with the throw as its reason', async () => {
|
|
141
|
+
const report = await runExperiment({
|
|
142
|
+
name: 'demo',
|
|
143
|
+
cases: [cases[0]],
|
|
144
|
+
scorers: [
|
|
145
|
+
customScorer('boom', () => {
|
|
146
|
+
throw new Error('scorer broke');
|
|
147
|
+
}),
|
|
148
|
+
],
|
|
149
|
+
run: async () => run(),
|
|
150
|
+
});
|
|
151
|
+
expect(report.cases[0]?.scores.boom?.score).toBe(0);
|
|
152
|
+
expect(report.cases[0]?.scores.boom?.reason).toContain('scorer broke');
|
|
153
|
+
});
|
|
154
|
+
it('preserves case order under concurrency', async () => {
|
|
155
|
+
const many = Array.from({ length: 8 }, (_, i) => ({
|
|
156
|
+
name: `case-${i}`,
|
|
157
|
+
input: i,
|
|
158
|
+
}));
|
|
159
|
+
const report = await runExperiment({
|
|
160
|
+
name: 'demo',
|
|
161
|
+
cases: many,
|
|
162
|
+
concurrency: 4,
|
|
163
|
+
scorers: [completionScorer()],
|
|
164
|
+
run: async (input) => {
|
|
165
|
+
await new Promise((r) => setTimeout(r, (8 - input) * 2));
|
|
166
|
+
return run();
|
|
167
|
+
},
|
|
168
|
+
});
|
|
169
|
+
expect(report.cases.map((c) => c.case)).toEqual(many.map((c) => c.name));
|
|
170
|
+
});
|
|
171
|
+
it('honours a per-case scorer override', async () => {
|
|
172
|
+
const report = await runExperiment({
|
|
173
|
+
name: 'demo',
|
|
174
|
+
cases: [{ name: 'special', input: null, scorers: [containsScorer('xyz')] }],
|
|
175
|
+
scorers: [completionScorer()],
|
|
176
|
+
run: async () => run({ output: 'no match here' }),
|
|
177
|
+
});
|
|
178
|
+
expect(Object.keys(report.cases[0]?.scores ?? {})).toEqual(['contains']);
|
|
179
|
+
});
|
|
180
|
+
it('respects a partial pass threshold', async () => {
|
|
181
|
+
const report = await runExperiment({
|
|
182
|
+
name: 'demo',
|
|
183
|
+
cases: [{ name: 'partial', input: null, expectedTools: ['read', 'edit'] }],
|
|
184
|
+
scorers: [trajectoryScorer()],
|
|
185
|
+
passThreshold: 0.5,
|
|
186
|
+
run: async () => run({ toolCalls: ['read'] }),
|
|
187
|
+
});
|
|
188
|
+
expect(report.passed).toBe(1);
|
|
189
|
+
});
|
|
190
|
+
});
|
|
191
|
+
describe('formatReport', () => {
|
|
192
|
+
it('prints scorer reasons for failures, not just numbers', async () => {
|
|
193
|
+
// A CI log that says "0.62" sends someone back to reproduce it by hand.
|
|
194
|
+
const report = await runExperiment({
|
|
195
|
+
name: 'demo',
|
|
196
|
+
cases: [{ name: 'bad', input: null, expectedTools: ['read', 'edit'] }],
|
|
197
|
+
scorers: [trajectoryScorer()],
|
|
198
|
+
run: async () => run({ toolCalls: ['bash'] }),
|
|
199
|
+
});
|
|
200
|
+
const text = formatReport(report);
|
|
201
|
+
expect(text).toContain('0/1 passed');
|
|
202
|
+
expect(text).toContain('✗ bad');
|
|
203
|
+
expect(text).toContain('trajectory:');
|
|
204
|
+
expect(text).toContain('expected read → edit');
|
|
205
|
+
});
|
|
206
|
+
});
|
|
207
|
+
//# sourceMappingURL=harness.test.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"harness.test.js","sourceRoot":"","sources":["../../../src/eval/__tests__/harness.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAE7C,OAAO,EAAE,YAAY,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAA;AAC9D,OAAO,EACN,gBAAgB,EAChB,cAAc,EACd,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,GAChB,MAAM,eAAe,CAAA;AAGtB;;;;;;;;GAQG;AAEH,SAAS,GAAG,CAAC,YAA8B,EAAE;IAC5C,OAAO;QACN,MAAM,EAAE,MAAM;QACd,KAAK,EAAE,EAAE;QACT,SAAS,EAAE,EAAE;QACb,UAAU,EAAE,UAAU;QACtB,WAAW,EAAE,CAAC;QACd,YAAY,EAAE,CAAC;QACf,UAAU,EAAE,CAAC;QACb,GAAG,SAAS;KACZ,CAAA;AACF,CAAC;AAED,MAAM,IAAI,GAAa,EAAE,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,IAAI,EAAE,CAAA;AAEjD,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE;IACjC,qEAAqE;IACrE,wEAAwE;IACxE,MAAM,KAAK,GAAG,KAAK,EAAE,aAAuB,EAAE,SAAmB,EAAE,EAAE,CACpE,MAAM,gBAAgB,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,SAAS,EAAE,CAAC,EAAE,EAAE,GAAG,IAAI,EAAE,aAAa,EAAE,CAAC,CAAA;IAE/E,EAAE,CAAC,8BAA8B,EAAE,KAAK,IAAI,EAAE;QAC7C,MAAM,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACxE,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mDAAmD,EAAE,KAAK,IAAI,EAAE;QAClE,sEAAsE;QACtE,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC,CAAA;QAChE,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;QACtC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC,CAAA;IAC1C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,2CAA2C,EAAE,KAAK,IAAI,EAAE;QAC1D,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAA;QAChE,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;QACtC,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACxC,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;IACpD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,yCAAyC,EAAE,KAAK,IAAI,EAAE;QACxD,MAAM,IAAI,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAA;QAC5D,MAAM,CAAC,IAAI,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACvC,MAAM,CAAC,IAAI,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;IAC7C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,2EAA2E,EAAE,KAAK,IAAI,EAAE;QAC1F,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,IAAI,EAAE,MAAM,CAAC,CAAC,CAAA;QACtE,MAAM,OAAO,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAA;QACvD,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,CAAA;IAC/C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mDAAmD,EAAE,KAAK,IAAI,EAAE;QAClE,MAAM,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC5C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,gDAAgD,EAAE,KAAK,IAAI,EAAE;QAC/D,MAAM,CAAC,GAAG,MAAM,KAAK,CAAC,CAAC,MAAM,CAAC,EAAE,EAAE,CAAC,CAAA;QACnC,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACvB,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,sBAAsB,CAAC,CAAA;IACnD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,+CAA+C,EAAE,KAAK,IAAI,EAAE;QAC9D,MAAM,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAClD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,wBAAwB,EAAE,KAAK,IAAI,EAAE;QACvC,MAAM,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC,CAAA;IAC3E,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE;IACjC,EAAE,CAAC,wBAAwB,EAAE,KAAK,IAAI,EAAE;QACvC,MAAM,CAAC,CAAC,MAAM,gBAAgB,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,UAAU,EAAE,UAAU,EAAE,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC9F,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,8BAA8B,EAAE,KAAK,IAAI,EAAE;QAC7C,MAAM,CAAC,GAAG,MAAM,gBAAgB,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,UAAU,EAAE,gBAAgB,EAAE,CAAC,EAAE,IAAI,CAAC,CAAA;QACrF,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACvB,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,gBAAgB,CAAC,CAAA;IAC7C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACrD,MAAM,CAAC,GAAG,MAAM,gBAAgB,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,EAAE,IAAI,CAAC,CAAA;QACtE,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACvB,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAA;IACnC,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE;IACjC,MAAM,SAAS,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,GAAG,CAAC,EAAE,KAAK,EAAE,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,CAAC,EAAE,CAAU,CAAC,EAAE,CAAC,CAAA;IAE/F,EAAE,CAAC,sBAAsB,EAAE,KAAK,IAAI,EAAE;QACrC,MAAM,CAAC,CAAC,MAAM,gBAAgB,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC5E,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,+DAA+D,EAAE,KAAK,IAAI,EAAE;QAC9E,kEAAkE;QAClE,8CAA8C;QAC9C,MAAM,MAAM,GAAG,CAAC,MAAM,gBAAgB,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAA;QAC1E,MAAM,MAAM,GAAG,CAAC,MAAM,gBAAgB,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC,EAAE,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAA;QAC3E,MAAM,CAAC,MAAM,CAAC,CAAC,eAAe,CAAC,MAAM,CAAC,CAAA;QACtC,MAAM,CAAC,MAAM,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;IAC/B,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,gBAAgB,EAAE,GAAG,EAAE;IAC/B,EAAE,CAAC,qDAAqD,EAAE,KAAK,IAAI,EAAE;QACpE,MAAM,CAAC,GAAG,MAAM,cAAc,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,MAAM,EAAE,YAAY,EAAE,CAAC,EAAE,IAAI,CAAC,CAAA;QAC1F,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;QACzB,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAA;IACnC,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,eAAe,EAAE,GAAG,EAAE;IAC9B,MAAM,KAAK,GAAe;QACzB,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,GAAG,EAAE,aAAa,EAAE,CAAC,MAAM,CAAC,EAAE;QACrD,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,GAAG,EAAE,aAAa,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE;KAC5D,CAAA;IAED,EAAE,CAAC,kCAAkC,EAAE,KAAK,IAAI,EAAE;QACjD,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK;YACL,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,GAAG,EAAE,KAAK,EAAE,KAAK,EAAE,EAAE,CAAC,GAAG,CAAC,EAAE,SAAS,EAAE,KAAK,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC;SAC7E,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;QACpC,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC,CAAC,WAAW,CAAC,GAAG,EAAE,CAAC,CAAC,CAAA;IACvD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0CAA0C,EAAE,KAAK,IAAI,EAAE;QACzD,qEAAqE;QACrE,eAAe;QACf,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK;YACL,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,GAAG,EAAE,KAAK,EAAE,KAAK,EAAE,EAAE;gBACpB,IAAI,KAAK,KAAK,GAAG;oBAAE,MAAM,IAAI,KAAK,CAAC,UAAU,CAAC,CAAA;gBAC9C,OAAO,GAAG,EAAE,CAAA;YACb,CAAC;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;QACpC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAA;QACnD,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;IAC3C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,4DAA4D,EAAE,KAAK,IAAI,EAAE;QAC3E,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,KAAK,CAAC,CAAC,CAAa,CAAC;YAC7B,OAAO,EAAE;gBACR,YAAY,CAAC,MAAM,EAAE,GAAG,EAAE;oBACzB,MAAM,IAAI,KAAK,CAAC,cAAc,CAAC,CAAA;gBAChC,CAAC,CAAC;aACF;YACD,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,GAAG,EAAE;SACtB,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACnD,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,SAAS,CAAC,cAAc,CAAC,CAAA;IACvE,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,wCAAwC,EAAE,KAAK,IAAI,EAAE;QACvD,MAAM,IAAI,GAAe,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;YAC7D,IAAI,EAAE,QAAQ,CAAC,EAAE;YACjB,KAAK,EAAE,CAAC;SACR,CAAC,CAAC,CAAA;QACH,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,IAAI;YACX,WAAW,EAAE,CAAC;YACd,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,GAAG,EAAE,KAAK,EAAE,KAAK,EAAE,EAAE;gBACpB,MAAM,IAAI,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,CAAC,EAAE,CAAC,CAAC,GAAI,KAAgB,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;gBACpE,OAAO,GAAG,EAAE,CAAA;YACb,CAAC;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAA;IACzE,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oCAAoC,EAAE,KAAK,IAAI,EAAE;QACnD,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,cAAc,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC;YAC3E,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,GAAG,CAAC,EAAE,MAAM,EAAE,eAAe,EAAE,CAAC;SACjD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,IAAI,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,UAAU,CAAC,CAAC,CAAA;IACzE,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mCAAmC,EAAE,KAAK,IAAI,EAAE;QAClD,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,IAAI,EAAE,aAAa,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC;YAC1E,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,aAAa,EAAE,GAAG;YAClB,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,GAAG,CAAC,EAAE,SAAS,EAAE,CAAC,MAAM,CAAC,EAAE,CAAC;SAC7C,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC9B,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE;IAC7B,EAAE,CAAC,sDAAsD,EAAE,KAAK,IAAI,EAAE;QACrE,wEAAwE;QACxE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,IAAI,EAAE,MAAM;YACZ,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,aAAa,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC;YACtE,OAAO,EAAE,CAAC,gBAAgB,EAAE,CAAC;YAC7B,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,GAAG,CAAC,EAAE,SAAS,EAAE,CAAC,MAAM,CAAC,EAAE,CAAC;SAC7C,CAAC,CAAA;QAEF,MAAM,IAAI,GAAG,YAAY,CAAC,MAAM,CAAC,CAAA;QACjC,MAAM,CAAC,IAAI,CAAC,CAAC,SAAS,CAAC,YAAY,CAAC,CAAA;QACpC,MAAM,CAAC,IAAI,CAAC,CAAC,SAAS,CAAC,OAAO,CAAC,CAAA;QAC/B,MAAM,CAAC,IAAI,CAAC,CAAC,SAAS,CAAC,aAAa,CAAC,CAAA;QACrC,MAAM,CAAC,IAAI,CAAC,CAAC,SAAS,CAAC,sBAAsB,CAAC,CAAA;IAC/C,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"judge.test.d.ts","sourceRoot":"","sources":["../../../src/eval/__tests__/judge.test.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,255 @@
|
|
|
1
|
+
import { describe, expect, it } from 'vitest';
|
|
2
|
+
import { runExperiment } from '../experiment.js';
|
|
3
|
+
import { judgeScorer } from '../judge.js';
|
|
4
|
+
/**
|
|
5
|
+
* Every other scorer is a pure function over the run, which is what makes
|
|
6
|
+
* them reproducible and what makes them unable to say whether an answer is
|
|
7
|
+
* GOOD. A judge closes that, and brings a failure mode none of the others
|
|
8
|
+
* have: it can fail to answer at all. A failed measurement scored zero
|
|
9
|
+
* reads exactly like a regression, so the distinction is what most of this
|
|
10
|
+
* file is about.
|
|
11
|
+
*/
|
|
12
|
+
function fakeProvider(reply, tokens = 42) {
|
|
13
|
+
const seen = [];
|
|
14
|
+
const provider = {
|
|
15
|
+
id: 'fake',
|
|
16
|
+
name: 'fake',
|
|
17
|
+
capabilities: {
|
|
18
|
+
supportsTools: false,
|
|
19
|
+
supportsStreaming: true,
|
|
20
|
+
supportsFunctionCalling: false,
|
|
21
|
+
},
|
|
22
|
+
seen,
|
|
23
|
+
async *chatStream(params) {
|
|
24
|
+
seen.push(params);
|
|
25
|
+
if (typeof reply !== 'string')
|
|
26
|
+
reply();
|
|
27
|
+
yield { id: 'j', delta: { content: reply } };
|
|
28
|
+
yield {
|
|
29
|
+
id: 'j',
|
|
30
|
+
delta: {},
|
|
31
|
+
finishReason: 'stop',
|
|
32
|
+
usage: {
|
|
33
|
+
promptTokens: tokens,
|
|
34
|
+
completionTokens: 0,
|
|
35
|
+
totalTokens: tokens,
|
|
36
|
+
cachedTokens: 0,
|
|
37
|
+
cacheWriteTokens: 0,
|
|
38
|
+
},
|
|
39
|
+
};
|
|
40
|
+
},
|
|
41
|
+
async listModels() {
|
|
42
|
+
return [];
|
|
43
|
+
},
|
|
44
|
+
async healthCheck() {
|
|
45
|
+
return true;
|
|
46
|
+
},
|
|
47
|
+
};
|
|
48
|
+
return provider;
|
|
49
|
+
}
|
|
50
|
+
const RUN = {
|
|
51
|
+
output: 'The capital of France is Paris.',
|
|
52
|
+
steps: [],
|
|
53
|
+
toolCalls: ['search'],
|
|
54
|
+
stopReason: 'end_turn',
|
|
55
|
+
totalTokens: 100,
|
|
56
|
+
totalCostUsd: 0,
|
|
57
|
+
durationMs: 10,
|
|
58
|
+
};
|
|
59
|
+
const CASE = { name: 'capital', input: 'What is the capital of France?' };
|
|
60
|
+
const RUBRIC = 'The answer names the correct city.';
|
|
61
|
+
describe('grading', () => {
|
|
62
|
+
it('normalises the grade onto 0..1 against the scale it asked for', async () => {
|
|
63
|
+
const scorer = judgeScorer({
|
|
64
|
+
provider: fakeProvider('{"grade": 3, "reason": "names the right city"}'),
|
|
65
|
+
model: 'm',
|
|
66
|
+
rubric: RUBRIC,
|
|
67
|
+
scale: 4,
|
|
68
|
+
});
|
|
69
|
+
const score = await scorer.score(RUN, CASE);
|
|
70
|
+
expect(score.score).toBe(0.75);
|
|
71
|
+
expect(score.reason).toBe('names the right city');
|
|
72
|
+
expect(score.details?.grade).toBe(3);
|
|
73
|
+
});
|
|
74
|
+
it('reads a verdict wrapped in prose or fences', async () => {
|
|
75
|
+
const scorer = judgeScorer({
|
|
76
|
+
provider: fakeProvider('Here is my grade:\n```json\n{"grade": 4, "reason": "correct"}\n```'),
|
|
77
|
+
model: 'm',
|
|
78
|
+
rubric: RUBRIC,
|
|
79
|
+
});
|
|
80
|
+
expect((await scorer.score(RUN, CASE)).score).toBe(1);
|
|
81
|
+
});
|
|
82
|
+
it('handles a brace inside the reason string', async () => {
|
|
83
|
+
const scorer = judgeScorer({
|
|
84
|
+
provider: fakeProvider('{"grade": 2, "reason": "mentions {city} placeholder"}'),
|
|
85
|
+
model: 'm',
|
|
86
|
+
rubric: RUBRIC,
|
|
87
|
+
scale: 4,
|
|
88
|
+
});
|
|
89
|
+
// A naive scan to the first `}` would cut the object in half here.
|
|
90
|
+
expect((await scorer.score(RUN, CASE)).score).toBe(0.5);
|
|
91
|
+
});
|
|
92
|
+
it('carries what the judging itself cost', async () => {
|
|
93
|
+
const scorer = judgeScorer({
|
|
94
|
+
provider: fakeProvider('{"grade": 4, "reason": "ok"}', 1234),
|
|
95
|
+
model: 'm',
|
|
96
|
+
rubric: RUBRIC,
|
|
97
|
+
});
|
|
98
|
+
// A judge is the most expensive scorer there is; a bill nobody can
|
|
99
|
+
// attribute is a bill nobody controls.
|
|
100
|
+
expect((await scorer.score(RUN, CASE)).details?.judgeTokens).toBe(1234);
|
|
101
|
+
});
|
|
102
|
+
it('grades the same run the same way twice', async () => {
|
|
103
|
+
const provider = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
104
|
+
await judgeScorer({ provider, model: 'm', rubric: RUBRIC }).score(RUN, CASE);
|
|
105
|
+
// Sampling noise is indistinguishable from a regression, so the
|
|
106
|
+
// judge does not sample.
|
|
107
|
+
expect(provider.seen[0]?.temperature).toBe(0);
|
|
108
|
+
});
|
|
109
|
+
});
|
|
110
|
+
describe('what the judge is shown', () => {
|
|
111
|
+
const promptOf = (provider) => String(provider.seen[0]?.messages[0]?.content);
|
|
112
|
+
it('shows the rubric, the task and the answer', async () => {
|
|
113
|
+
const provider = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
114
|
+
await judgeScorer({ provider, model: 'm', rubric: RUBRIC }).score(RUN, CASE);
|
|
115
|
+
const prompt = promptOf(provider);
|
|
116
|
+
expect(prompt).toContain(RUBRIC);
|
|
117
|
+
expect(prompt).toContain('What is the capital of France?');
|
|
118
|
+
expect(prompt).toContain('The capital of France is Paris.');
|
|
119
|
+
});
|
|
120
|
+
it('withholds the trajectory unless asked', async () => {
|
|
121
|
+
const off = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
122
|
+
await judgeScorer({ provider: off, model: 'm', rubric: RUBRIC }).score(RUN, CASE);
|
|
123
|
+
expect(promptOf(off)).not.toContain('TOOLS CALLED');
|
|
124
|
+
const on = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
125
|
+
await judgeScorer({
|
|
126
|
+
provider: on,
|
|
127
|
+
model: 'm',
|
|
128
|
+
rubric: RUBRIC,
|
|
129
|
+
includeTrajectory: true,
|
|
130
|
+
}).score(RUN, CASE);
|
|
131
|
+
expect(promptOf(on)).toContain('search');
|
|
132
|
+
});
|
|
133
|
+
it('says when it truncated the answer', async () => {
|
|
134
|
+
const provider = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
135
|
+
await judgeScorer({
|
|
136
|
+
provider,
|
|
137
|
+
model: 'm',
|
|
138
|
+
rubric: RUBRIC,
|
|
139
|
+
maxOutputChars: 10,
|
|
140
|
+
}).score({ ...RUN, output: 'x'.repeat(500) }, CASE);
|
|
141
|
+
// A judge shown a silently cut answer marks it down for stopping
|
|
142
|
+
// mid-sentence, which scores our truncation rather than the run.
|
|
143
|
+
expect(promptOf(provider)).toContain('cut at 10 characters by the harness');
|
|
144
|
+
});
|
|
145
|
+
it('offers the reference answer when the case has one', async () => {
|
|
146
|
+
const provider = fakeProvider('{"grade": 4, "reason": "ok"}');
|
|
147
|
+
await judgeScorer({ provider, model: 'm', rubric: RUBRIC }).score(RUN, {
|
|
148
|
+
...CASE,
|
|
149
|
+
expected: 'Paris',
|
|
150
|
+
});
|
|
151
|
+
expect(promptOf(provider)).toContain('Paris');
|
|
152
|
+
});
|
|
153
|
+
});
|
|
154
|
+
describe('a judge that cannot judge', () => {
|
|
155
|
+
const failing = (reply) => judgeScorer({ provider: fakeProvider(reply), model: 'm', rubric: RUBRIC, scale: 4 });
|
|
156
|
+
it('throws on a reply with no JSON at all', async () => {
|
|
157
|
+
await expect(failing('I think it was pretty good!').score(RUN, CASE)).rejects.toThrow(/no JSON/);
|
|
158
|
+
});
|
|
159
|
+
it('throws on a grade outside the scale it was given', async () => {
|
|
160
|
+
// Clamping would turn a judge that misread the scale into a
|
|
161
|
+
// confident score, and a judge that misread the scale did not apply
|
|
162
|
+
// the rubric either.
|
|
163
|
+
await expect(failing('{"grade": 9, "reason": "great"}').score(RUN, CASE)).rejects.toThrow(/outside the 0\.\.4 scale/);
|
|
164
|
+
});
|
|
165
|
+
it('throws on a non-numeric grade', async () => {
|
|
166
|
+
await expect(failing('{"grade": "good", "reason": "x"}').score(RUN, CASE)).rejects.toThrow(/no numeric grade/);
|
|
167
|
+
});
|
|
168
|
+
it('throws on an unterminated object', async () => {
|
|
169
|
+
await expect(failing('{"grade": 3, "reason": "cut off').score(RUN, CASE)).rejects.toThrow(/unterminated/);
|
|
170
|
+
});
|
|
171
|
+
it('refuses to exist without a rubric', () => {
|
|
172
|
+
expect(() => judgeScorer({ provider: fakeProvider('{}'), model: 'm', rubric: ' ' })).toThrow(/rubric is required/);
|
|
173
|
+
});
|
|
174
|
+
it('refuses a nonsense scale', () => {
|
|
175
|
+
expect(() => judgeScorer({ provider: fakeProvider('{}'), model: 'm', rubric: RUBRIC, scale: 0 })).toThrow(/positive integer/);
|
|
176
|
+
});
|
|
177
|
+
});
|
|
178
|
+
describe('a broken judge is not a bad run', () => {
|
|
179
|
+
const dataset = {
|
|
180
|
+
name: 'suite',
|
|
181
|
+
cases: [CASE],
|
|
182
|
+
run: async () => RUN,
|
|
183
|
+
};
|
|
184
|
+
it('reports the case as inconclusive rather than failed', async () => {
|
|
185
|
+
const report = await runExperiment({
|
|
186
|
+
...dataset,
|
|
187
|
+
scorers: [
|
|
188
|
+
judgeScorer({
|
|
189
|
+
provider: fakeProvider(() => {
|
|
190
|
+
throw new Error('429 rate limited');
|
|
191
|
+
}),
|
|
192
|
+
model: 'm',
|
|
193
|
+
rubric: RUBRIC,
|
|
194
|
+
}),
|
|
195
|
+
],
|
|
196
|
+
});
|
|
197
|
+
expect(report.inconclusive).toBe(1);
|
|
198
|
+
expect(report.failed).toBe(0);
|
|
199
|
+
expect(report.passed).toBe(0);
|
|
200
|
+
expect(report.cases[0]?.scores.judge?.unavailable).toBe(true);
|
|
201
|
+
expect(report.cases[0]?.scores.judge?.reason).toContain('429');
|
|
202
|
+
});
|
|
203
|
+
it('keeps an unavailable score out of the mean entirely', async () => {
|
|
204
|
+
const report = await runExperiment({
|
|
205
|
+
...dataset,
|
|
206
|
+
scorers: [
|
|
207
|
+
{ name: 'always-one', score: () => ({ score: 1, reason: 'fine' }) },
|
|
208
|
+
judgeScorer({
|
|
209
|
+
provider: fakeProvider(() => {
|
|
210
|
+
throw new Error('down');
|
|
211
|
+
}),
|
|
212
|
+
model: 'm',
|
|
213
|
+
rubric: RUBRIC,
|
|
214
|
+
}),
|
|
215
|
+
],
|
|
216
|
+
});
|
|
217
|
+
// Not 0.5. A measurement that never happened is not a zero, and
|
|
218
|
+
// averaging it in would report a regression the harness caused.
|
|
219
|
+
expect(report.cases[0]?.mean).toBe(1);
|
|
220
|
+
expect(report.cases[0]?.status).toBe('passed');
|
|
221
|
+
expect(report.mean).toBe(1);
|
|
222
|
+
});
|
|
223
|
+
it('omits a scorer that was never available from the per-dimension means', async () => {
|
|
224
|
+
const report = await runExperiment({
|
|
225
|
+
...dataset,
|
|
226
|
+
scorers: [
|
|
227
|
+
judgeScorer({
|
|
228
|
+
provider: fakeProvider(() => {
|
|
229
|
+
throw new Error('down');
|
|
230
|
+
}),
|
|
231
|
+
model: 'm',
|
|
232
|
+
rubric: RUBRIC,
|
|
233
|
+
}),
|
|
234
|
+
],
|
|
235
|
+
});
|
|
236
|
+
// A dimension with no measurements is not a dimension that scored 0.
|
|
237
|
+
expect(report.byScorer.judge).toBeUndefined();
|
|
238
|
+
});
|
|
239
|
+
it('still fails a case the judge actually judged badly', async () => {
|
|
240
|
+
const report = await runExperiment({
|
|
241
|
+
...dataset,
|
|
242
|
+
scorers: [
|
|
243
|
+
judgeScorer({
|
|
244
|
+
provider: fakeProvider('{"grade": 0, "reason": "names the wrong city"}'),
|
|
245
|
+
model: 'm',
|
|
246
|
+
rubric: RUBRIC,
|
|
247
|
+
}),
|
|
248
|
+
],
|
|
249
|
+
});
|
|
250
|
+
expect(report.failed).toBe(1);
|
|
251
|
+
expect(report.inconclusive).toBe(0);
|
|
252
|
+
expect(report.cases[0]?.scores.judge?.unavailable).toBeUndefined();
|
|
253
|
+
});
|
|
254
|
+
});
|
|
255
|
+
//# sourceMappingURL=judge.test.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"judge.test.js","sourceRoot":"","sources":["../../../src/eval/__tests__/judge.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAI7C,OAAO,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAA;AAChD,OAAO,EAAE,WAAW,EAAE,MAAM,aAAa,CAAA;AAGzC;;;;;;;GAOG;AAEH,SAAS,YAAY,CAAC,KAA6B,EAAE,MAAM,GAAG,EAAE;IAC/D,MAAM,IAAI,GAA2B,EAAE,CAAA;IACvC,MAAM,QAAQ,GAAmD;QAChE,EAAE,EAAE,MAAM;QACV,IAAI,EAAE,MAAM;QACZ,YAAY,EAAE;YACb,aAAa,EAAE,KAAK;YACpB,iBAAiB,EAAE,IAAI;YACvB,uBAAuB,EAAE,KAAK;SAC9B;QACD,IAAI;QACJ,KAAK,CAAC,CAAC,UAAU,CAAC,MAA4B;YAC7C,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;YACjB,IAAI,OAAO,KAAK,KAAK,QAAQ;gBAAE,KAAK,EAAE,CAAA;YACtC,MAAM,EAAE,EAAE,EAAE,GAAG,EAAE,KAAK,EAAE,EAAE,OAAO,EAAE,KAAe,EAAE,EAAE,CAAA;YACtD,MAAM;gBACL,EAAE,EAAE,GAAG;gBACP,KAAK,EAAE,EAAE;gBACT,YAAY,EAAE,MAAM;gBACpB,KAAK,EAAE;oBACN,YAAY,EAAE,MAAM;oBACpB,gBAAgB,EAAE,CAAC;oBACnB,WAAW,EAAE,MAAM;oBACnB,YAAY,EAAE,CAAC;oBACf,gBAAgB,EAAE,CAAC;iBACnB;aACD,CAAA;QACF,CAAC;QACD,KAAK,CAAC,UAAU;YACf,OAAO,EAAE,CAAA;QACV,CAAC;QACD,KAAK,CAAC,WAAW;YAChB,OAAO,IAAI,CAAA;QACZ,CAAC;KACD,CAAA;IACD,OAAO,QAAQ,CAAA;AAChB,CAAC;AAED,MAAM,GAAG,GAAY;IACpB,MAAM,EAAE,iCAAiC;IACzC,KAAK,EAAE,EAAE;IACT,SAAS,EAAE,CAAC,QAAQ,CAAC;IACrB,UAAU,EAAE,UAAU;IACtB,WAAW,EAAE,GAAG;IAChB,YAAY,EAAE,CAAC;IACf,UAAU,EAAE,EAAE;CACd,CAAA;AAED,MAAM,IAAI,GAAa,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,gCAAgC,EAAE,CAAA;AAEnF,MAAM,MAAM,GAAG,oCAAoC,CAAA;AAEnD,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE;IACxB,EAAE,CAAC,+DAA+D,EAAE,KAAK,IAAI,EAAE;QAC9E,MAAM,MAAM,GAAG,WAAW,CAAC;YAC1B,QAAQ,EAAE,YAAY,CAAC,gDAAgD,CAAC;YACxE,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;YACd,KAAK,EAAE,CAAC;SACR,CAAC,CAAA;QACF,MAAM,KAAK,GAAG,MAAM,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAA;QAE3C,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;QAC9B,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,sBAAsB,CAAC,CAAA;QACjD,MAAM,CAAC,KAAK,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACrC,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,4CAA4C,EAAE,KAAK,IAAI,EAAE;QAC3D,MAAM,MAAM,GAAG,WAAW,CAAC;YAC1B,QAAQ,EAAE,YAAY,CAAC,oEAAoE,CAAC;YAC5F,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;SACd,CAAC,CAAA;QACF,MAAM,CAAC,CAAC,MAAM,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACtD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0CAA0C,EAAE,KAAK,IAAI,EAAE;QACzD,MAAM,MAAM,GAAG,WAAW,CAAC;YAC1B,QAAQ,EAAE,YAAY,CAAC,uDAAuD,CAAC;YAC/E,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;YACd,KAAK,EAAE,CAAC;SACR,CAAC,CAAA;QACF,mEAAmE;QACnE,MAAM,CAAC,CAAC,MAAM,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;IACxD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACrD,MAAM,MAAM,GAAG,WAAW,CAAC;YAC1B,QAAQ,EAAE,YAAY,CAAC,8BAA8B,EAAE,IAAI,CAAC;YAC5D,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;SACd,CAAC,CAAA;QACF,mEAAmE;QACnE,uCAAuC;QACvC,MAAM,CAAC,CAAC,MAAM,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,EAAE,WAAW,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;IACxE,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,wCAAwC,EAAE,KAAK,IAAI,EAAE;QACvD,MAAM,QAAQ,GAAG,YAAY,CAAC,8BAA8B,CAE3D,CAAA;QACD,MAAM,WAAW,CAAC,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAA;QAC5E,gEAAgE;QAChE,yBAAyB;QACzB,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,WAAW,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC9C,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,yBAAyB,EAAE,GAAG,EAAE;IACxC,MAAM,QAAQ,GAAG,CAAC,QAAqB,EAAE,EAAE,CAC1C,MAAM,CAAE,QAAwD,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,QAAQ,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAA;IAEhG,EAAE,CAAC,2CAA2C,EAAE,KAAK,IAAI,EAAE;QAC1D,MAAM,QAAQ,GAAG,YAAY,CAAC,8BAA8B,CAAC,CAAA;QAC7D,MAAM,WAAW,CAAC,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAA;QAE5E,MAAM,MAAM,GAAG,QAAQ,CAAC,QAAQ,CAAC,CAAA;QACjC,MAAM,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAA;QAChC,MAAM,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,gCAAgC,CAAC,CAAA;QAC1D,MAAM,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,iCAAiC,CAAC,CAAA;IAC5D,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,uCAAuC,EAAE,KAAK,IAAI,EAAE;QACtD,MAAM,GAAG,GAAG,YAAY,CAAC,8BAA8B,CAAC,CAAA;QACxD,MAAM,WAAW,CAAC,EAAE,QAAQ,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAA;QACjF,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,cAAc,CAAC,CAAA;QAEnD,MAAM,EAAE,GAAG,YAAY,CAAC,8BAA8B,CAAC,CAAA;QACvD,MAAM,WAAW,CAAC;YACjB,QAAQ,EAAE,EAAE;YACZ,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;YACd,iBAAiB,EAAE,IAAI;SACvB,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAA;QACnB,MAAM,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAA;IACzC,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mCAAmC,EAAE,KAAK,IAAI,EAAE;QAClD,MAAM,QAAQ,GAAG,YAAY,CAAC,8BAA8B,CAAC,CAAA;QAC7D,MAAM,WAAW,CAAC;YACjB,QAAQ;YACR,KAAK,EAAE,GAAG;YACV,MAAM,EAAE,MAAM;YACd,cAAc,EAAE,EAAE;SAClB,CAAC,CAAC,KAAK,CAAC,EAAE,GAAG,GAAG,EAAE,MAAM,EAAE,GAAG,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,EAAE,IAAI,CAAC,CAAA;QAEnD,iEAAiE;QACjE,iEAAiE;QACjE,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC,CAAC,SAAS,CAAC,qCAAqC,CAAC,CAAA;IAC5E,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mDAAmD,EAAE,KAAK,IAAI,EAAE;QAClE,MAAM,QAAQ,GAAG,YAAY,CAAC,8BAA8B,CAAC,CAAA;QAC7D,MAAM,WAAW,CAAC,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE;YACtE,GAAG,IAAI;YACP,QAAQ,EAAE,OAAO;SACjB,CAAC,CAAA;QACF,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC,CAAC,SAAS,CAAC,OAAO,CAAC,CAAA;IAC9C,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,2BAA2B,EAAE,GAAG,EAAE;IAC1C,MAAM,OAAO,GAAG,CAAC,KAAa,EAAE,EAAE,CACjC,WAAW,CAAC,EAAE,QAAQ,EAAE,YAAY,CAAC,KAAK,CAAC,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAA;IAErF,EAAE,CAAC,uCAAuC,EAAE,KAAK,IAAI,EAAE;QACtD,MAAM,MAAM,CAAC,OAAO,CAAC,6BAA6B,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,SAAS,CAAC,CAAA;IACjG,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,kDAAkD,EAAE,KAAK,IAAI,EAAE;QACjE,4DAA4D;QAC5D,oEAAoE;QACpE,qBAAqB;QACrB,MAAM,MAAM,CAAC,OAAO,CAAC,iCAAiC,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CACxF,0BAA0B,CAC1B,CAAA;IACF,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,+BAA+B,EAAE,KAAK,IAAI,EAAE;QAC9C,MAAM,MAAM,CAAC,OAAO,CAAC,kCAAkC,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CACzF,kBAAkB,CAClB,CAAA;IACF,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,kCAAkC,EAAE,KAAK,IAAI,EAAE;QACjD,MAAM,MAAM,CAAC,OAAO,CAAC,iCAAiC,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CACxF,cAAc,CACd,CAAA;IACF,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mCAAmC,EAAE,GAAG,EAAE;QAC5C,MAAM,CAAC,GAAG,EAAE,CAAC,WAAW,CAAC,EAAE,QAAQ,EAAE,YAAY,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,OAAO,CAC7F,oBAAoB,CACpB,CAAA;IACF,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0BAA0B,EAAE,GAAG,EAAE;QACnC,MAAM,CAAC,GAAG,EAAE,CACX,WAAW,CAAC,EAAE,QAAQ,EAAE,YAAY,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CACnF,CAAC,OAAO,CAAC,kBAAkB,CAAC,CAAA;IAC9B,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA;AAEF,QAAQ,CAAC,iCAAiC,EAAE,GAAG,EAAE;IAChD,MAAM,OAAO,GAAG;QACf,IAAI,EAAE,OAAO;QACb,KAAK,EAAE,CAAC,IAAI,CAAC;QACb,GAAG,EAAE,KAAK,IAAsB,EAAE,CAAC,GAAG;KACtC,CAAA;IAED,EAAE,CAAC,qDAAqD,EAAE,KAAK,IAAI,EAAE;QACpE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,GAAG,OAAO;YACV,OAAO,EAAE;gBACR,WAAW,CAAC;oBACX,QAAQ,EAAE,YAAY,CAAC,GAAG,EAAE;wBAC3B,MAAM,IAAI,KAAK,CAAC,kBAAkB,CAAC,CAAA;oBACpC,CAAC,CAAC;oBACF,KAAK,EAAE,GAAG;oBACV,MAAM,EAAE,MAAM;iBACd,CAAC;aACF;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,YAAY,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACnC,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,EAAE,WAAW,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;QAC7D,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,SAAS,CAAC,KAAK,CAAC,CAAA;IAC/D,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,qDAAqD,EAAE,KAAK,IAAI,EAAE;QACpE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,GAAG,OAAO;YACV,OAAO,EAAE;gBACR,EAAE,IAAI,EAAE,YAAY,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,EAAE;gBACnE,WAAW,CAAC;oBACX,QAAQ,EAAE,YAAY,CAAC,GAAG,EAAE;wBAC3B,MAAM,IAAI,KAAK,CAAC,MAAM,CAAC,CAAA;oBACxB,CAAC,CAAC;oBACF,KAAK,EAAE,GAAG;oBACV,MAAM,EAAE,MAAM;iBACd,CAAC;aACF;SACD,CAAC,CAAA;QAEF,gEAAgE;QAChE,gEAAgE;QAChE,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACrC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;QAC9C,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IAC5B,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,sEAAsE,EAAE,KAAK,IAAI,EAAE;QACrF,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,GAAG,OAAO;YACV,OAAO,EAAE;gBACR,WAAW,CAAC;oBACX,QAAQ,EAAE,YAAY,CAAC,GAAG,EAAE;wBAC3B,MAAM,IAAI,KAAK,CAAC,MAAM,CAAC,CAAA;oBACxB,CAAC,CAAC;oBACF,KAAK,EAAE,GAAG;oBACV,MAAM,EAAE,MAAM;iBACd,CAAC;aACF;SACD,CAAC,CAAA;QAEF,qEAAqE;QACrE,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,aAAa,EAAE,CAAA;IAC9C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oDAAoD,EAAE,KAAK,IAAI,EAAE;QACnE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC;YAClC,GAAG,OAAO;YACV,OAAO,EAAE;gBACR,WAAW,CAAC;oBACX,QAAQ,EAAE,YAAY,CAAC,gDAAgD,CAAC;oBACxE,KAAK,EAAE,GAAG;oBACV,MAAM,EAAE,MAAM;iBACd,CAAC;aACF;SACD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,CAAC,MAAM,CAAC,YAAY,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACnC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,EAAE,WAAW,CAAC,CAAC,aAAa,EAAE,CAAA;IACnE,CAAC,CAAC,CAAA;AACH,CAAC,CAAC,CAAA"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"scorer-gates.test.d.ts","sourceRoot":"","sources":["../../../src/eval/__tests__/scorer-gates.test.ts"],"names":[],"mappings":""}
|