cognitive-core 0.2.5 → 0.3.1
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/LICENSE +21 -0
- package/README.md +88 -143
- package/dist/agents/mock-provider.d.ts.map +1 -1
- package/dist/agents/mock-provider.js.map +1 -1
- package/dist/agents/types.d.ts.map +1 -1
- package/dist/agents/types.js +5 -4
- package/dist/agents/types.js.map +1 -1
- package/dist/atlas.d.ts +309 -97
- package/dist/atlas.d.ts.map +1 -1
- package/dist/atlas.js +743 -277
- package/dist/atlas.js.map +1 -1
- package/dist/bin/cli-utils.d.ts +25 -2
- package/dist/bin/cli-utils.d.ts.map +1 -1
- package/dist/bin/cli-utils.js +79 -9
- package/dist/bin/cli-utils.js.map +1 -1
- package/dist/bin/cognitive-core.js +32 -0
- package/dist/bin/cognitive-core.js.map +1 -1
- package/dist/bin/commands/dream.d.ts +17 -0
- package/dist/bin/commands/dream.d.ts.map +1 -0
- package/dist/bin/commands/dream.js +168 -0
- package/dist/bin/commands/dream.js.map +1 -0
- package/dist/bin/commands/eval.d.ts +181 -0
- package/dist/bin/commands/eval.d.ts.map +1 -0
- package/dist/bin/commands/eval.js +1582 -0
- package/dist/bin/commands/eval.js.map +1 -0
- package/dist/bin/commands/kb.js +0 -8
- package/dist/bin/commands/kb.js.map +1 -1
- package/dist/bin/commands/learn.js +12 -4
- package/dist/bin/commands/learn.js.map +1 -1
- package/dist/bin/commands/legacy.d.ts.map +1 -1
- package/dist/bin/commands/legacy.js +1 -1
- package/dist/bin/commands/legacy.js.map +1 -1
- package/dist/bin/commands/run.d.ts.map +1 -1
- package/dist/bin/commands/run.js +10 -9
- package/dist/bin/commands/run.js.map +1 -1
- package/dist/bin/commands/sessions.d.ts.map +1 -1
- package/dist/bin/commands/sessions.js.map +1 -1
- package/dist/bin/commands/skills.d.ts.map +1 -1
- package/dist/bin/commands/skills.js +283 -1
- package/dist/bin/commands/skills.js.map +1 -1
- package/dist/embeddings/bm25.d.ts.map +1 -1
- package/dist/embeddings/bm25.js +1 -2
- package/dist/embeddings/bm25.js.map +1 -1
- package/dist/embeddings/index.d.ts +3 -3
- package/dist/embeddings/index.d.ts.map +1 -1
- package/dist/embeddings/index.js +3 -3
- package/dist/embeddings/index.js.map +1 -1
- package/dist/embeddings/inverted-index.js +62 -7
- package/dist/embeddings/inverted-index.js.map +1 -1
- package/dist/embeddings/manager.d.ts.map +1 -1
- package/dist/embeddings/manager.js +6 -5
- package/dist/embeddings/manager.js.map +1 -1
- package/dist/embeddings/provider.d.ts +96 -5
- package/dist/embeddings/provider.d.ts.map +1 -1
- package/dist/embeddings/provider.js +171 -48
- package/dist/embeddings/provider.js.map +1 -1
- package/dist/embeddings/vector-store.d.ts.map +1 -1
- package/dist/embeddings/vector-store.js +3 -1
- package/dist/embeddings/vector-store.js.map +1 -1
- package/dist/eval/ablation-validate.d.ts +12 -0
- package/dist/eval/ablation-validate.d.ts.map +1 -0
- package/dist/eval/ablation-validate.js +78 -0
- package/dist/eval/ablation-validate.js.map +1 -0
- package/dist/eval/ablation.d.ts +94 -0
- package/dist/eval/ablation.d.ts.map +1 -0
- package/dist/eval/ablation.js +161 -0
- package/dist/eval/ablation.js.map +1 -0
- package/dist/eval/ci-smoke.d.ts +11 -0
- package/dist/eval/ci-smoke.d.ts.map +1 -0
- package/dist/eval/ci-smoke.js +49 -0
- package/dist/eval/ci-smoke.js.map +1 -0
- package/dist/eval/diagnostics.d.ts +214 -0
- package/dist/eval/diagnostics.d.ts.map +1 -0
- package/dist/eval/diagnostics.js +107 -0
- package/dist/eval/diagnostics.js.map +1 -0
- package/dist/eval/index.d.ts +27 -0
- package/dist/eval/index.d.ts.map +1 -0
- package/dist/eval/index.js +27 -0
- package/dist/eval/index.js.map +1 -0
- package/dist/eval/live/alfworld.d.ts +248 -0
- package/dist/eval/live/alfworld.d.ts.map +1 -0
- package/dist/eval/live/alfworld.js +637 -0
- package/dist/eval/live/alfworld.js.map +1 -0
- package/dist/eval/live/alfworld_h1.d.ts +16 -0
- package/dist/eval/live/alfworld_h1.d.ts.map +1 -0
- package/dist/eval/live/alfworld_h1.js +75 -0
- package/dist/eval/live/alfworld_h1.js.map +1 -0
- package/dist/eval/live/alfworld_memory.d.ts +38 -0
- package/dist/eval/live/alfworld_memory.d.ts.map +1 -0
- package/dist/eval/live/alfworld_memory.js +184 -0
- package/dist/eval/live/alfworld_memory.js.map +1 -0
- package/dist/eval/live/alfworld_strategy.d.ts +65 -0
- package/dist/eval/live/alfworld_strategy.d.ts.map +1 -0
- package/dist/eval/live/alfworld_strategy.js +89 -0
- package/dist/eval/live/alfworld_strategy.js.map +1 -0
- package/dist/eval/live/azure-openai-llm-provider.d.ts +41 -0
- package/dist/eval/live/azure-openai-llm-provider.d.ts.map +1 -0
- package/dist/eval/live/azure-openai-llm-provider.js +92 -0
- package/dist/eval/live/azure-openai-llm-provider.js.map +1 -0
- package/dist/eval/live/bridges/skillflow_e2b_bridge.d.ts +3 -0
- package/dist/eval/live/bridges/skillflow_e2b_bridge.d.ts.map +1 -0
- package/dist/eval/live/bridges/skillflow_e2b_bridge.js +12 -0
- package/dist/eval/live/bridges/skillflow_e2b_bridge.js.map +1 -0
- package/dist/eval/live/claude-cli-llm-provider.d.ts +53 -0
- package/dist/eval/live/claude-cli-llm-provider.d.ts.map +1 -0
- package/dist/eval/live/claude-cli-llm-provider.js +71 -0
- package/dist/eval/live/claude-cli-llm-provider.js.map +1 -0
- package/dist/eval/live/claude-cli.d.ts +96 -0
- package/dist/eval/live/claude-cli.d.ts.map +1 -0
- package/dist/eval/live/claude-cli.js +314 -0
- package/dist/eval/live/claude-cli.js.map +1 -0
- package/dist/eval/live/claude-delegate.d.ts +41 -0
- package/dist/eval/live/claude-delegate.d.ts.map +1 -0
- package/dist/eval/live/claude-delegate.js +59 -0
- package/dist/eval/live/claude-delegate.js.map +1 -0
- package/dist/eval/live/claude-workspace-backend.d.ts +24 -0
- package/dist/eval/live/claude-workspace-backend.d.ts.map +1 -0
- package/dist/eval/live/claude-workspace-backend.js +76 -0
- package/dist/eval/live/claude-workspace-backend.js.map +1 -0
- package/dist/eval/live/clbench.d.ts +64 -0
- package/dist/eval/live/clbench.d.ts.map +1 -0
- package/dist/eval/live/clbench.js +291 -0
- package/dist/eval/live/clbench.js.map +1 -0
- package/dist/eval/live/cold-reuse.d.ts +21 -0
- package/dist/eval/live/cold-reuse.d.ts.map +1 -0
- package/dist/eval/live/cold-reuse.js +102 -0
- package/dist/eval/live/cold-reuse.js.map +1 -0
- package/dist/eval/live/difficulty-registry.d.ts +68 -0
- package/dist/eval/live/difficulty-registry.d.ts.map +1 -0
- package/dist/eval/live/difficulty-registry.js +169 -0
- package/dist/eval/live/difficulty-registry.js.map +1 -0
- package/dist/eval/live/docker.d.ts +52 -0
- package/dist/eval/live/docker.d.ts.map +1 -0
- package/dist/eval/live/docker.js +164 -0
- package/dist/eval/live/docker.js.map +1 -0
- package/dist/eval/live/env-bridge.d.ts +68 -0
- package/dist/eval/live/env-bridge.d.ts.map +1 -0
- package/dist/eval/live/env-bridge.js +188 -0
- package/dist/eval/live/env-bridge.js.map +1 -0
- package/dist/eval/live/gsm8k.d.ts +44 -0
- package/dist/eval/live/gsm8k.d.ts.map +1 -0
- package/dist/eval/live/gsm8k.js +142 -0
- package/dist/eval/live/gsm8k.js.map +1 -0
- package/dist/eval/live/index.d.ts +38 -0
- package/dist/eval/live/index.d.ts.map +1 -0
- package/dist/eval/live/index.js +38 -0
- package/dist/eval/live/index.js.map +1 -0
- package/dist/eval/live/lifelongagent.d.ts +64 -0
- package/dist/eval/live/lifelongagent.d.ts.map +1 -0
- package/dist/eval/live/lifelongagent.js +216 -0
- package/dist/eval/live/lifelongagent.js.map +1 -0
- package/dist/eval/live/lift-registry.d.ts +128 -0
- package/dist/eval/live/lift-registry.d.ts.map +1 -0
- package/dist/eval/live/lift-registry.js +147 -0
- package/dist/eval/live/lift-registry.js.map +1 -0
- package/dist/eval/live/live-benchmark.d.ts +84 -0
- package/dist/eval/live/live-benchmark.d.ts.map +1 -0
- package/dist/eval/live/live-benchmark.js +2 -0
- package/dist/eval/live/live-benchmark.js.map +1 -0
- package/dist/eval/live/live-runner.d.ts +609 -0
- package/dist/eval/live/live-runner.d.ts.map +1 -0
- package/dist/eval/live/live-runner.js +1409 -0
- package/dist/eval/live/live-runner.js.map +1 -0
- package/dist/eval/live/refine.d.ts +88 -0
- package/dist/eval/live/refine.d.ts.map +1 -0
- package/dist/eval/live/refine.js +135 -0
- package/dist/eval/live/refine.js.map +1 -0
- package/dist/eval/live/registry-store.d.ts +109 -0
- package/dist/eval/live/registry-store.d.ts.map +1 -0
- package/dist/eval/live/registry-store.js +272 -0
- package/dist/eval/live/registry-store.js.map +1 -0
- package/dist/eval/live/run-checkpoint.d.ts +55 -0
- package/dist/eval/live/run-checkpoint.d.ts.map +1 -0
- package/dist/eval/live/run-checkpoint.js +86 -0
- package/dist/eval/live/run-checkpoint.js.map +1 -0
- package/dist/eval/live/run-resume.d.ts +22 -0
- package/dist/eval/live/run-resume.d.ts.map +1 -0
- package/dist/eval/live/run-resume.js +70 -0
- package/dist/eval/live/run-resume.js.map +1 -0
- package/dist/eval/live/skill-injector.d.ts +36 -0
- package/dist/eval/live/skill-injector.d.ts.map +1 -0
- package/dist/eval/live/skill-injector.js +81 -0
- package/dist/eval/live/skill-injector.js.map +1 -0
- package/dist/eval/live/skillflow-embedded-workbook-bundle.d.ts +3 -0
- package/dist/eval/live/skillflow-embedded-workbook-bundle.d.ts.map +1 -0
- package/dist/eval/live/skillflow-embedded-workbook-bundle.js +391 -0
- package/dist/eval/live/skillflow-embedded-workbook-bundle.js.map +1 -0
- package/dist/eval/live/skillflow.d.ts +154 -0
- package/dist/eval/live/skillflow.d.ts.map +1 -0
- package/dist/eval/live/skillflow.js +612 -0
- package/dist/eval/live/skillflow.js.map +1 -0
- package/dist/eval/live/skilllearnbench.d.ts +43 -0
- package/dist/eval/live/skilllearnbench.d.ts.map +1 -0
- package/dist/eval/live/skilllearnbench.js +154 -0
- package/dist/eval/live/skilllearnbench.js.map +1 -0
- package/dist/eval/live/skillopt_v1.d.ts +96 -0
- package/dist/eval/live/skillopt_v1.d.ts.map +1 -0
- package/dist/eval/live/skillopt_v1.js +0 -0
- package/dist/eval/live/skillopt_v1.js.map +1 -0
- package/dist/eval/live/skillopt_v1_causal.d.ts +26 -0
- package/dist/eval/live/skillopt_v1_causal.d.ts.map +1 -0
- package/dist/eval/live/skillopt_v1_causal.js +106 -0
- package/dist/eval/live/skillopt_v1_causal.js.map +1 -0
- package/dist/eval/live/skillopt_v1_memory.d.ts +53 -0
- package/dist/eval/live/skillopt_v1_memory.d.ts.map +1 -0
- package/dist/eval/live/skillopt_v1_memory.js +147 -0
- package/dist/eval/live/skillopt_v1_memory.js.map +1 -0
- package/dist/eval/live/skillopt_v1_rules.d.ts +39 -0
- package/dist/eval/live/skillopt_v1_rules.d.ts.map +1 -0
- package/dist/eval/live/skillopt_v1_rules.js +91 -0
- package/dist/eval/live/skillopt_v1_rules.js.map +1 -0
- package/dist/eval/live/skillsbench.d.ts +138 -0
- package/dist/eval/live/skillsbench.d.ts.map +1 -0
- package/dist/eval/live/skillsbench.js +626 -0
- package/dist/eval/live/skillsbench.js.map +1 -0
- package/dist/eval/live/stopping.d.ts +69 -0
- package/dist/eval/live/stopping.d.ts.map +1 -0
- package/dist/eval/live/stopping.js +213 -0
- package/dist/eval/live/stopping.js.map +1 -0
- package/dist/eval/live/trace.d.ts +20 -0
- package/dist/eval/live/trace.d.ts.map +1 -0
- package/dist/eval/live/trace.js +41 -0
- package/dist/eval/live/trace.js.map +1 -0
- package/dist/eval/mechanism-trace-adapter.d.ts +8 -0
- package/dist/eval/mechanism-trace-adapter.d.ts.map +1 -0
- package/dist/eval/mechanism-trace-adapter.js +118 -0
- package/dist/eval/mechanism-trace-adapter.js.map +1 -0
- package/dist/eval/memory-ab.d.ts +61 -0
- package/dist/eval/memory-ab.d.ts.map +1 -0
- package/dist/eval/memory-ab.js +161 -0
- package/dist/eval/memory-ab.js.map +1 -0
- package/dist/eval/memory-mock.d.ts +54 -0
- package/dist/eval/memory-mock.d.ts.map +1 -0
- package/dist/eval/memory-mock.js +129 -0
- package/dist/eval/memory-mock.js.map +1 -0
- package/dist/eval/memory-qa/index.d.ts +13 -0
- package/dist/eval/memory-qa/index.d.ts.map +1 -0
- package/dist/eval/memory-qa/index.js +13 -0
- package/dist/eval/memory-qa/index.js.map +1 -0
- package/dist/eval/memory-qa/iterative.d.ts +80 -0
- package/dist/eval/memory-qa/iterative.d.ts.map +1 -0
- package/dist/eval/memory-qa/iterative.js +107 -0
- package/dist/eval/memory-qa/iterative.js.map +1 -0
- package/dist/eval/memory-qa/loaders.d.ts +11 -0
- package/dist/eval/memory-qa/loaders.d.ts.map +1 -0
- package/dist/eval/memory-qa/loaders.js +21 -0
- package/dist/eval/memory-qa/loaders.js.map +1 -0
- package/dist/eval/memory-qa/metrics.d.ts +3 -0
- package/dist/eval/memory-qa/metrics.d.ts.map +1 -0
- package/dist/eval/memory-qa/metrics.js +2 -0
- package/dist/eval/memory-qa/metrics.js.map +1 -0
- package/dist/eval/memory-qa/qa.d.ts +186 -0
- package/dist/eval/memory-qa/qa.d.ts.map +1 -0
- package/dist/eval/memory-qa/qa.js +424 -0
- package/dist/eval/memory-qa/qa.js.map +1 -0
- package/dist/eval/memory-qa/retrieval.d.ts +33 -0
- package/dist/eval/memory-qa/retrieval.d.ts.map +1 -0
- package/dist/eval/memory-qa/retrieval.js +44 -0
- package/dist/eval/memory-qa/retrieval.js.map +1 -0
- package/dist/eval/memory-qa/serving.d.ts +60 -0
- package/dist/eval/memory-qa/serving.d.ts.map +1 -0
- package/dist/eval/memory-qa/serving.js +267 -0
- package/dist/eval/memory-qa/serving.js.map +1 -0
- package/dist/eval/memory-qa/types.d.ts +2 -0
- package/dist/eval/memory-qa/types.d.ts.map +1 -0
- package/dist/eval/memory-qa/types.js +2 -0
- package/dist/eval/memory-qa/types.js.map +1 -0
- package/dist/eval/metrics.d.ts +120 -0
- package/dist/eval/metrics.d.ts.map +1 -0
- package/dist/eval/metrics.js +361 -0
- package/dist/eval/metrics.js.map +1 -0
- package/dist/eval/oracle-smoke.d.ts +75 -0
- package/dist/eval/oracle-smoke.d.ts.map +1 -0
- package/dist/eval/oracle-smoke.js +143 -0
- package/dist/eval/oracle-smoke.js.map +1 -0
- package/dist/eval/playbook-calibration-report.d.ts +40 -0
- package/dist/eval/playbook-calibration-report.d.ts.map +1 -0
- package/dist/eval/playbook-calibration-report.js +246 -0
- package/dist/eval/playbook-calibration-report.js.map +1 -0
- package/dist/eval/playbook-candidate-decisions-report.d.ts +36 -0
- package/dist/eval/playbook-candidate-decisions-report.d.ts.map +1 -0
- package/dist/eval/playbook-candidate-decisions-report.js +201 -0
- package/dist/eval/playbook-candidate-decisions-report.js.map +1 -0
- package/dist/eval/playbook-usage-report.d.ts +7 -0
- package/dist/eval/playbook-usage-report.d.ts.map +1 -0
- package/dist/eval/playbook-usage-report.js +222 -0
- package/dist/eval/playbook-usage-report.js.map +1 -0
- package/dist/eval/report.d.ts +28 -0
- package/dist/eval/report.d.ts.map +1 -0
- package/dist/eval/report.js +567 -0
- package/dist/eval/report.js.map +1 -0
- package/dist/eval/retrieval/corpus.d.ts +41 -0
- package/dist/eval/retrieval/corpus.d.ts.map +1 -0
- package/dist/eval/retrieval/corpus.js +158 -0
- package/dist/eval/retrieval/corpus.js.map +1 -0
- package/dist/eval/retrieval/index.d.ts +12 -0
- package/dist/eval/retrieval/index.d.ts.map +1 -0
- package/dist/eval/retrieval/index.js +14 -0
- package/dist/eval/retrieval/index.js.map +1 -0
- package/dist/eval/retrieval/knowledge-corpus.d.ts +91 -0
- package/dist/eval/retrieval/knowledge-corpus.d.ts.map +1 -0
- package/dist/eval/retrieval/knowledge-corpus.js +268 -0
- package/dist/eval/retrieval/knowledge-corpus.js.map +1 -0
- package/dist/eval/retrieval/knowledge-quality.d.ts +91 -0
- package/dist/eval/retrieval/knowledge-quality.d.ts.map +1 -0
- package/dist/eval/retrieval/knowledge-quality.js +247 -0
- package/dist/eval/retrieval/knowledge-quality.js.map +1 -0
- package/dist/eval/retrieval/ladder.d.ts +69 -0
- package/dist/eval/retrieval/ladder.d.ts.map +1 -0
- package/dist/eval/retrieval/ladder.js +80 -0
- package/dist/eval/retrieval/ladder.js.map +1 -0
- package/dist/eval/retrieval/quality.d.ts +39 -0
- package/dist/eval/retrieval/quality.d.ts.map +1 -0
- package/dist/eval/retrieval/quality.js +75 -0
- package/dist/eval/retrieval/quality.js.map +1 -0
- package/dist/eval/rng.d.ts +25 -0
- package/dist/eval/rng.d.ts.map +1 -0
- package/dist/eval/rng.js +43 -0
- package/dist/eval/rng.js.map +1 -0
- package/dist/eval/runner.d.ts +65 -0
- package/dist/eval/runner.d.ts.map +1 -0
- package/dist/eval/runner.js +216 -0
- package/dist/eval/runner.js.map +1 -0
- package/dist/eval/skill-bundle.d.ts +18 -0
- package/dist/eval/skill-bundle.d.ts.map +1 -0
- package/dist/eval/skill-bundle.js +38 -0
- package/dist/eval/skill-bundle.js.map +1 -0
- package/dist/eval/swarmkit-report.d.ts +21 -0
- package/dist/eval/swarmkit-report.d.ts.map +1 -0
- package/dist/eval/swarmkit-report.js +93 -0
- package/dist/eval/swarmkit-report.js.map +1 -0
- package/dist/eval/task-stream.d.ts +22 -0
- package/dist/eval/task-stream.d.ts.map +1 -0
- package/dist/eval/task-stream.js +99 -0
- package/dist/eval/task-stream.js.map +1 -0
- package/dist/eval/types.d.ts +173 -0
- package/dist/eval/types.d.ts.map +1 -0
- package/dist/eval/types.js +11 -0
- package/dist/eval/types.js.map +1 -0
- package/dist/experimental/causal-store.d.ts +97 -0
- package/dist/experimental/causal-store.d.ts.map +1 -0
- package/dist/experimental/causal-store.js +213 -0
- package/dist/experimental/causal-store.js.map +1 -0
- package/dist/experimental/graph-layers/base.d.ts +29 -0
- package/dist/experimental/graph-layers/base.d.ts.map +1 -0
- package/dist/experimental/graph-layers/base.js +143 -0
- package/dist/experimental/graph-layers/base.js.map +1 -0
- package/dist/experimental/graph-layers/causal.d.ts +14 -0
- package/dist/experimental/graph-layers/causal.d.ts.map +1 -0
- package/dist/experimental/graph-layers/causal.js +14 -0
- package/dist/experimental/graph-layers/causal.js.map +1 -0
- package/dist/experimental/graph-layers/entity.d.ts +14 -0
- package/dist/experimental/graph-layers/entity.d.ts.map +1 -0
- package/dist/experimental/graph-layers/entity.js +14 -0
- package/dist/experimental/graph-layers/entity.js.map +1 -0
- package/dist/experimental/graph-layers/index.d.ts +6 -0
- package/dist/experimental/graph-layers/index.d.ts.map +1 -0
- package/dist/experimental/graph-layers/index.js +6 -0
- package/dist/experimental/graph-layers/index.js.map +1 -0
- package/dist/experimental/graph-layers/semantic.d.ts +14 -0
- package/dist/experimental/graph-layers/semantic.d.ts.map +1 -0
- package/dist/experimental/graph-layers/semantic.js +14 -0
- package/dist/experimental/graph-layers/semantic.js.map +1 -0
- package/dist/experimental/graph-layers/temporal.d.ts +14 -0
- package/dist/experimental/graph-layers/temporal.d.ts.map +1 -0
- package/dist/experimental/graph-layers/temporal.js +14 -0
- package/dist/experimental/graph-layers/temporal.js.map +1 -0
- package/dist/experimental/healing-orchestrator.d.ts +182 -0
- package/dist/experimental/healing-orchestrator.d.ts.map +1 -0
- package/dist/experimental/healing-orchestrator.js +250 -0
- package/dist/experimental/healing-orchestrator.js.map +1 -0
- package/dist/experimental/knowledge-graph-types.d.ts +148 -0
- package/dist/experimental/knowledge-graph-types.d.ts.map +1 -0
- package/dist/experimental/knowledge-graph-types.js +40 -0
- package/dist/experimental/knowledge-graph-types.js.map +1 -0
- package/dist/experimental/knowledge-graph.d.ts +41 -0
- package/dist/experimental/knowledge-graph.d.ts.map +1 -0
- package/dist/experimental/knowledge-graph.js +270 -0
- package/dist/experimental/knowledge-graph.js.map +1 -0
- package/dist/experimental/meta-reflection.d.ts +47 -0
- package/dist/experimental/meta-reflection.d.ts.map +1 -0
- package/dist/experimental/meta-reflection.js +143 -0
- package/dist/experimental/meta-reflection.js.map +1 -0
- package/dist/experimental/moe-gate.d.ts +134 -0
- package/dist/experimental/moe-gate.d.ts.map +1 -0
- package/dist/experimental/moe-gate.js +240 -0
- package/dist/experimental/moe-gate.js.map +1 -0
- package/dist/experimental/reasoning-bank.d.ts +130 -0
- package/dist/experimental/reasoning-bank.d.ts.map +1 -0
- package/dist/experimental/reasoning-bank.js +427 -0
- package/dist/experimental/reasoning-bank.js.map +1 -0
- package/dist/experimental/team/team-config.d.ts +127 -0
- package/dist/experimental/team/team-config.d.ts.map +1 -0
- package/dist/experimental/team/team-config.js +48 -0
- package/dist/experimental/team/team-config.js.map +1 -0
- package/dist/experimental/team/team-experience.d.ts +300 -0
- package/dist/experimental/team/team-experience.d.ts.map +1 -0
- package/dist/experimental/team/team-experience.js +355 -0
- package/dist/experimental/team/team-experience.js.map +1 -0
- package/dist/experimental/team/team-ingester.d.ts +152 -0
- package/dist/experimental/team/team-ingester.d.ts.map +1 -0
- package/dist/experimental/team/team-ingester.js +330 -0
- package/dist/experimental/team/team-ingester.js.map +1 -0
- package/dist/experimental/team/team-meta-learner.d.ts +51 -0
- package/dist/experimental/team/team-meta-learner.d.ts.map +1 -0
- package/dist/experimental/team/team-meta-learner.js +404 -0
- package/dist/experimental/team/team-meta-learner.js.map +1 -0
- package/dist/experimental/team/team-meta.d.ts +160 -0
- package/dist/experimental/team/team-meta.d.ts.map +1 -0
- package/dist/experimental/team/team-meta.js +42 -0
- package/dist/experimental/team/team-meta.js.map +1 -0
- package/dist/experimental/team/team-pipeline.d.ts +77 -0
- package/dist/experimental/team/team-pipeline.d.ts.map +1 -0
- package/dist/experimental/team/team-pipeline.js +262 -0
- package/dist/experimental/team/team-pipeline.js.map +1 -0
- package/dist/experimental/team/team-playbook-extraction.d.ts +45 -0
- package/dist/experimental/team/team-playbook-extraction.d.ts.map +1 -0
- package/dist/experimental/team/team-playbook-extraction.js +526 -0
- package/dist/experimental/team/team-playbook-extraction.js.map +1 -0
- package/dist/experimental/team/team-playbook.d.ts +276 -0
- package/dist/experimental/team/team-playbook.d.ts.map +1 -0
- package/dist/experimental/team/team-playbook.js +85 -0
- package/dist/experimental/team/team-playbook.js.map +1 -0
- package/dist/experimental/team/team-router.d.ts +98 -0
- package/dist/experimental/team/team-router.d.ts.map +1 -0
- package/dist/experimental/team/team-router.js +312 -0
- package/dist/experimental/team/team-router.js.map +1 -0
- package/dist/experimental/team/team-skill-library.d.ts +180 -0
- package/dist/experimental/team/team-skill-library.d.ts.map +1 -0
- package/dist/experimental/team/team-skill-library.js +385 -0
- package/dist/experimental/team/team-skill-library.js.map +1 -0
- package/dist/experimental/team/team-trajectory-analysis.d.ts +19 -0
- package/dist/experimental/team/team-trajectory-analysis.d.ts.map +1 -0
- package/dist/experimental/team/team-trajectory-analysis.js +442 -0
- package/dist/experimental/team/team-trajectory-analysis.js.map +1 -0
- package/dist/experimental/team/team-trajectory.d.ts +305 -0
- package/dist/experimental/team/team-trajectory.d.ts.map +1 -0
- package/dist/experimental/team/team-trajectory.js +315 -0
- package/dist/experimental/team/team-trajectory.js.map +1 -0
- package/dist/factory.d.ts +2 -3
- package/dist/factory.d.ts.map +1 -1
- package/dist/factory.js +12 -11
- package/dist/factory.js.map +1 -1
- package/dist/index.d.ts +9 -11
- package/dist/index.d.ts.map +1 -1
- package/dist/index.js +26 -32
- package/dist/index.js.map +1 -1
- package/dist/learning/analyzer.d.ts +26 -36
- package/dist/learning/analyzer.d.ts.map +1 -1
- package/dist/learning/analyzer.js +50 -11
- package/dist/learning/analyzer.js.map +1 -1
- package/dist/learning/compound-bootstrap/external-acquisition.d.ts +63 -0
- package/dist/learning/compound-bootstrap/external-acquisition.d.ts.map +1 -0
- package/dist/learning/compound-bootstrap/external-acquisition.js +255 -0
- package/dist/learning/compound-bootstrap/external-acquisition.js.map +1 -0
- package/dist/learning/compound-bootstrap/index.d.ts +3 -0
- package/dist/learning/compound-bootstrap/index.d.ts.map +1 -0
- package/dist/learning/compound-bootstrap/index.js +3 -0
- package/dist/learning/compound-bootstrap/index.js.map +1 -0
- package/dist/learning/compound-bootstrap/workspace-analyzer.d.ts +33 -0
- package/dist/learning/compound-bootstrap/workspace-analyzer.d.ts.map +1 -0
- package/dist/learning/compound-bootstrap/workspace-analyzer.js +379 -0
- package/dist/learning/compound-bootstrap/workspace-analyzer.js.map +1 -0
- package/dist/learning/consolidation-coordinator.d.ts +49 -0
- package/dist/learning/consolidation-coordinator.d.ts.map +1 -0
- package/dist/learning/consolidation-coordinator.js +75 -0
- package/dist/learning/consolidation-coordinator.js.map +1 -0
- package/dist/learning/convention-distiller.d.ts +84 -0
- package/dist/learning/convention-distiller.d.ts.map +1 -0
- package/dist/learning/convention-distiller.js +268 -0
- package/dist/learning/convention-distiller.js.map +1 -0
- package/dist/learning/distiller.d.ts +82 -0
- package/dist/learning/distiller.d.ts.map +1 -0
- package/dist/learning/distiller.js +353 -0
- package/dist/learning/distiller.js.map +1 -0
- package/dist/learning/effectiveness.d.ts +12 -2
- package/dist/learning/effectiveness.d.ts.map +1 -1
- package/dist/learning/effectiveness.js +18 -14
- package/dist/learning/effectiveness.js.map +1 -1
- package/dist/learning/escalation.d.ts +89 -0
- package/dist/learning/escalation.d.ts.map +1 -0
- package/dist/learning/escalation.js +168 -0
- package/dist/learning/escalation.js.map +1 -0
- package/dist/learning/external-exposure.d.ts +57 -0
- package/dist/learning/external-exposure.d.ts.map +1 -0
- package/dist/learning/external-exposure.js +95 -0
- package/dist/learning/external-exposure.js.map +1 -0
- package/dist/learning/gepa-proposer.d.ts +116 -0
- package/dist/learning/gepa-proposer.d.ts.map +1 -0
- package/dist/learning/gepa-proposer.js +258 -0
- package/dist/learning/gepa-proposer.js.map +1 -0
- package/dist/learning/index.d.ts +21 -8
- package/dist/learning/index.d.ts.map +1 -1
- package/dist/learning/index.js +28 -10
- package/dist/learning/index.js.map +1 -1
- package/dist/learning/instant-loop.d.ts +46 -16
- package/dist/learning/instant-loop.d.ts.map +1 -1
- package/dist/learning/instant-loop.js +114 -111
- package/dist/learning/instant-loop.js.map +1 -1
- package/dist/learning/knowledge-extractor.d.ts.map +1 -1
- package/dist/learning/knowledge-extractor.js +1 -2
- package/dist/learning/knowledge-extractor.js.map +1 -1
- package/dist/learning/llm-extractor.d.ts +88 -0
- package/dist/learning/llm-extractor.d.ts.map +1 -0
- package/dist/learning/llm-extractor.js +372 -0
- package/dist/learning/llm-extractor.js.map +1 -0
- package/dist/learning/loop-coordinator.d.ts +61 -0
- package/dist/learning/loop-coordinator.d.ts.map +1 -0
- package/dist/learning/loop-coordinator.js +96 -0
- package/dist/learning/loop-coordinator.js.map +1 -0
- package/dist/learning/maintenance-scheduler.d.ts +15 -4
- package/dist/learning/maintenance-scheduler.d.ts.map +1 -1
- package/dist/learning/maintenance-scheduler.js +19 -16
- package/dist/learning/maintenance-scheduler.js.map +1 -1
- package/dist/learning/mechanism-trace.d.ts +196 -0
- package/dist/learning/mechanism-trace.d.ts.map +1 -0
- package/dist/learning/mechanism-trace.js +81 -0
- package/dist/learning/mechanism-trace.js.map +1 -0
- package/dist/learning/merge-proposer.d.ts +102 -0
- package/dist/learning/merge-proposer.d.ts.map +1 -0
- package/dist/learning/merge-proposer.js +253 -0
- package/dist/learning/merge-proposer.js.map +1 -0
- package/dist/learning/meta-learner.d.ts +9 -7
- package/dist/learning/meta-learner.d.ts.map +1 -1
- package/dist/learning/meta-learner.js +16 -45
- package/dist/learning/meta-learner.js.map +1 -1
- package/dist/learning/mutation-gate.d.ts +218 -0
- package/dist/learning/mutation-gate.d.ts.map +1 -0
- package/dist/learning/mutation-gate.js +390 -0
- package/dist/learning/mutation-gate.js.map +1 -0
- package/dist/learning/oracle-signal.d.ts +67 -0
- package/dist/learning/oracle-signal.d.ts.map +1 -0
- package/dist/learning/oracle-signal.js +94 -0
- package/dist/learning/oracle-signal.js.map +1 -0
- package/dist/learning/pipeline.d.ts +116 -0
- package/dist/learning/pipeline.d.ts.map +1 -0
- package/dist/learning/pipeline.js +288 -0
- package/dist/learning/pipeline.js.map +1 -0
- package/dist/learning/playbook-extractor.d.ts +8 -0
- package/dist/learning/playbook-extractor.d.ts.map +1 -1
- package/dist/learning/playbook-extractor.js +119 -34
- package/dist/learning/playbook-extractor.js.map +1 -1
- package/dist/learning/playbook-optimizer.d.ts +37 -0
- package/dist/learning/playbook-optimizer.d.ts.map +1 -0
- package/dist/learning/playbook-optimizer.js +125 -0
- package/dist/learning/playbook-optimizer.js.map +1 -0
- package/dist/learning/reflective-proposer.d.ts +173 -0
- package/dist/learning/reflective-proposer.d.ts.map +1 -0
- package/dist/learning/reflective-proposer.js +354 -0
- package/dist/learning/reflective-proposer.js.map +1 -0
- package/dist/learning/reflexion-generator.d.ts.map +1 -1
- package/dist/learning/reflexion-generator.js +4 -3
- package/dist/learning/reflexion-generator.js.map +1 -1
- package/dist/learning/replay-executor.d.ts +64 -0
- package/dist/learning/replay-executor.d.ts.map +1 -0
- package/dist/learning/replay-executor.js +92 -0
- package/dist/learning/replay-executor.js.map +1 -0
- package/dist/learning/replay-types.d.ts +150 -0
- package/dist/learning/replay-types.d.ts.map +1 -0
- package/dist/learning/replay-types.js +207 -0
- package/dist/learning/replay-types.js.map +1 -0
- package/dist/learning/replay.d.ts +166 -0
- package/dist/learning/replay.d.ts.map +1 -0
- package/dist/learning/replay.js +228 -0
- package/dist/learning/replay.js.map +1 -0
- package/dist/learning/seed-report.d.ts +88 -0
- package/dist/learning/seed-report.d.ts.map +1 -0
- package/dist/learning/seed-report.js +120 -0
- package/dist/learning/seed-report.js.map +1 -0
- package/dist/learning/trajectory-sources/dataclaw.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/dataclaw.js +6 -18
- package/dist/learning/trajectory-sources/dataclaw.js.map +1 -1
- package/dist/learning/trajectory-sources/entire.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/entire.js +30 -9
- package/dist/learning/trajectory-sources/entire.js.map +1 -1
- package/dist/learning/trajectory-sources/file.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/file.js.map +1 -1
- package/dist/learning/trajectory-sources/huggingface.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/huggingface.js +7 -7
- package/dist/learning/trajectory-sources/huggingface.js.map +1 -1
- package/dist/learning/trajectory-sources/in-memory.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/in-memory.js.map +1 -1
- package/dist/learning/trajectory-sources/index.d.ts +3 -4
- package/dist/learning/trajectory-sources/index.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/index.js +3 -4
- package/dist/learning/trajectory-sources/index.js.map +1 -1
- package/dist/learning/trajectory-sources/pipeline.d.ts +4 -3
- package/dist/learning/trajectory-sources/pipeline.d.ts.map +1 -1
- package/dist/learning/trajectory-sources/pipeline.js +4 -3
- package/dist/learning/trajectory-sources/pipeline.js.map +1 -1
- package/dist/learning/unified-pipeline.d.ts +486 -61
- package/dist/learning/unified-pipeline.d.ts.map +1 -1
- package/dist/learning/unified-pipeline.js +1501 -232
- package/dist/learning/unified-pipeline.js.map +1 -1
- package/dist/learning/usage-inference.d.ts.map +1 -1
- package/dist/learning/usage-inference.js +4 -9
- package/dist/learning/usage-inference.js.map +1 -1
- package/dist/learning/workspace-snapshot.d.ts +146 -0
- package/dist/learning/workspace-snapshot.d.ts.map +1 -0
- package/dist/learning/workspace-snapshot.js +269 -0
- package/dist/learning/workspace-snapshot.js.map +1 -0
- package/dist/mcp/playbook-server.d.ts +61 -2
- package/dist/mcp/playbook-server.d.ts.map +1 -1
- package/dist/mcp/playbook-server.js +179 -9
- package/dist/mcp/playbook-server.js.map +1 -1
- package/dist/memory/coherence.d.ts.map +1 -1
- package/dist/memory/coherence.js +5 -7
- package/dist/memory/coherence.js.map +1 -1
- package/dist/memory/curated-loader.d.ts +24 -0
- package/dist/memory/curated-loader.d.ts.map +1 -1
- package/dist/memory/curated-loader.js +22 -5
- package/dist/memory/curated-loader.js.map +1 -1
- package/dist/memory/evolution/llm-evolver.d.ts +36 -0
- package/dist/memory/evolution/llm-evolver.d.ts.map +1 -0
- package/dist/memory/evolution/llm-evolver.js +106 -0
- package/dist/memory/evolution/llm-evolver.js.map +1 -0
- package/dist/memory/experience.d.ts +34 -13
- package/dist/memory/experience.d.ts.map +1 -1
- package/dist/memory/experience.js +59 -48
- package/dist/memory/experience.js.map +1 -1
- package/dist/memory/extraction/conversation-extractor.d.ts +62 -0
- package/dist/memory/extraction/conversation-extractor.d.ts.map +1 -0
- package/dist/memory/extraction/conversation-extractor.js +169 -0
- package/dist/memory/extraction/conversation-extractor.js.map +1 -0
- package/dist/memory/filesystem-cascade.d.ts +116 -0
- package/dist/memory/filesystem-cascade.d.ts.map +1 -0
- package/dist/memory/filesystem-cascade.js +221 -0
- package/dist/memory/filesystem-cascade.js.map +1 -0
- package/dist/memory/index.d.ts +14 -10
- package/dist/memory/index.d.ts.map +1 -1
- package/dist/memory/index.js +15 -12
- package/dist/memory/index.js.map +1 -1
- package/dist/memory/knowledge-bank.d.ts +145 -29
- package/dist/memory/knowledge-bank.d.ts.map +1 -1
- package/dist/memory/knowledge-bank.js +511 -200
- package/dist/memory/knowledge-bank.js.map +1 -1
- package/dist/memory/knowledge-defrag.d.ts +32 -9
- package/dist/memory/knowledge-defrag.d.ts.map +1 -1
- package/dist/memory/knowledge-defrag.js +57 -49
- package/dist/memory/knowledge-defrag.js.map +1 -1
- package/dist/memory/meta-playbook-bank.d.ts +24 -0
- package/dist/memory/meta-playbook-bank.d.ts.map +1 -0
- package/dist/memory/meta-playbook-bank.js +135 -0
- package/dist/memory/meta-playbook-bank.js.map +1 -0
- package/dist/memory/meta.d.ts.map +1 -1
- package/dist/memory/meta.js +33 -14
- package/dist/memory/meta.js.map +1 -1
- package/dist/memory/playbook-applicability.d.ts +43 -0
- package/dist/memory/playbook-applicability.d.ts.map +1 -0
- package/dist/memory/playbook-applicability.js +155 -0
- package/dist/memory/playbook-applicability.js.map +1 -0
- package/dist/memory/playbook-incorporation.d.ts +24 -0
- package/dist/memory/playbook-incorporation.d.ts.map +1 -0
- package/dist/memory/playbook-incorporation.js +41 -0
- package/dist/memory/playbook-incorporation.js.map +1 -0
- package/dist/memory/playbook.d.ts +444 -7
- package/dist/memory/playbook.d.ts.map +1 -1
- package/dist/memory/playbook.js +1231 -50
- package/dist/memory/playbook.js.map +1 -1
- package/dist/memory/procedural-skill.d.ts +142 -0
- package/dist/memory/procedural-skill.d.ts.map +1 -0
- package/dist/memory/procedural-skill.js +846 -0
- package/dist/memory/procedural-skill.js.map +1 -0
- package/dist/memory/search-provider.d.ts +8 -3
- package/dist/memory/search-provider.d.ts.map +1 -1
- package/dist/memory/search-providers/embedding.d.ts +64 -0
- package/dist/memory/search-providers/embedding.d.ts.map +1 -0
- package/dist/memory/search-providers/embedding.js +96 -0
- package/dist/memory/search-providers/embedding.js.map +1 -0
- package/dist/memory/search-providers/hybrid.d.ts +74 -0
- package/dist/memory/search-providers/hybrid.d.ts.map +1 -0
- package/dist/memory/search-providers/hybrid.js +94 -0
- package/dist/memory/search-providers/hybrid.js.map +1 -0
- package/dist/memory/search-providers/index.d.ts +3 -0
- package/dist/memory/search-providers/index.d.ts.map +1 -1
- package/dist/memory/search-providers/index.js +3 -0
- package/dist/memory/search-providers/index.js.map +1 -1
- package/dist/memory/search-providers/keyword-expanding.d.ts +22 -0
- package/dist/memory/search-providers/keyword-expanding.d.ts.map +1 -0
- package/dist/memory/search-providers/keyword-expanding.js +35 -0
- package/dist/memory/search-providers/keyword-expanding.js.map +1 -0
- package/dist/memory/search-providers/minimem.d.ts +16 -4
- package/dist/memory/search-providers/minimem.d.ts.map +1 -1
- package/dist/memory/search-providers/minimem.js +14 -3
- package/dist/memory/search-providers/minimem.js.map +1 -1
- package/dist/memory/search-providers/text-similarity.d.ts +2 -4
- package/dist/memory/search-providers/text-similarity.d.ts.map +1 -1
- package/dist/memory/search-providers/text-similarity.js.map +1 -1
- package/dist/memory/seeds/meta-playbook-seeds.d.ts +27 -0
- package/dist/memory/seeds/meta-playbook-seeds.d.ts.map +1 -0
- package/dist/memory/seeds/meta-playbook-seeds.js +560 -0
- package/dist/memory/seeds/meta-playbook-seeds.js.map +1 -0
- package/dist/memory/seeds/meta-strategy-seeds.d.ts +23 -0
- package/dist/memory/seeds/meta-strategy-seeds.d.ts.map +1 -0
- package/dist/memory/seeds/meta-strategy-seeds.js +82 -0
- package/dist/memory/seeds/meta-strategy-seeds.js.map +1 -0
- package/dist/memory/seeds/seed-loader.d.ts +34 -0
- package/dist/memory/seeds/seed-loader.d.ts.map +1 -0
- package/dist/memory/seeds/seed-loader.js +60 -0
- package/dist/memory/seeds/seed-loader.js.map +1 -0
- package/dist/memory/skill-exporter.d.ts.map +1 -1
- package/dist/memory/skill-exporter.js +1 -2
- package/dist/memory/skill-exporter.js.map +1 -1
- package/dist/memory/source-resolver.d.ts.map +1 -1
- package/dist/memory/source-resolver.js +7 -8
- package/dist/memory/source-resolver.js.map +1 -1
- package/dist/memory/system.d.ts +119 -14
- package/dist/memory/system.d.ts.map +1 -1
- package/dist/memory/system.js +416 -60
- package/dist/memory/system.js.map +1 -1
- package/dist/memory/temporal-compressor.d.ts +1 -1
- package/dist/memory/temporal-compressor.d.ts.map +1 -1
- package/dist/memory/temporal-compressor.js +26 -16
- package/dist/memory/temporal-compressor.js.map +1 -1
- package/dist/persistence/index.d.ts +2 -2
- package/dist/persistence/index.d.ts.map +1 -1
- package/dist/persistence/index.js +1 -1
- package/dist/persistence/index.js.map +1 -1
- package/dist/persistence/migrator.d.ts +34 -1
- package/dist/persistence/migrator.d.ts.map +1 -1
- package/dist/persistence/migrator.js +163 -33
- package/dist/persistence/migrator.js.map +1 -1
- package/dist/persistence/sqlite-persistence.d.ts +412 -1
- package/dist/persistence/sqlite-persistence.d.ts.map +1 -1
- package/dist/persistence/sqlite-persistence.js +1038 -47
- package/dist/persistence/sqlite-persistence.js.map +1 -1
- package/dist/runtime/backends/acp-protocol.d.ts.map +1 -1
- package/dist/runtime/backends/acp-protocol.js +2 -1
- package/dist/runtime/backends/acp-protocol.js.map +1 -1
- package/dist/runtime/backends/acp.d.ts.map +1 -1
- package/dist/runtime/backends/acp.js +5 -4
- package/dist/runtime/backends/acp.js.map +1 -1
- package/dist/runtime/backends/index.d.ts +2 -2
- package/dist/runtime/backends/index.d.ts.map +1 -1
- package/dist/runtime/backends/index.js +2 -2
- package/dist/runtime/backends/index.js.map +1 -1
- package/dist/runtime/backends/macro-agent.d.ts +17 -1
- package/dist/runtime/backends/macro-agent.d.ts.map +1 -1
- package/dist/runtime/backends/macro-agent.js +1 -1
- package/dist/runtime/backends/macro-agent.js.map +1 -1
- package/dist/runtime/backends/mock.d.ts.map +1 -1
- package/dist/runtime/backends/mock.js.map +1 -1
- package/dist/runtime/backends/subprocess.d.ts.map +1 -1
- package/dist/runtime/backends/subprocess.js +5 -4
- package/dist/runtime/backends/subprocess.js.map +1 -1
- package/dist/runtime/compute-provider.d.ts +2 -18
- package/dist/runtime/compute-provider.d.ts.map +1 -1
- package/dist/runtime/compute-provider.js +8 -8
- package/dist/runtime/compute-provider.js.map +1 -1
- package/dist/runtime/flows/learning.d.ts +2 -1
- package/dist/runtime/flows/learning.d.ts.map +1 -1
- package/dist/runtime/flows/learning.js +4 -5
- package/dist/runtime/flows/learning.js.map +1 -1
- package/dist/runtime/flows/validation.d.ts.map +1 -1
- package/dist/runtime/flows/validation.js +2 -5
- package/dist/runtime/flows/validation.js.map +1 -1
- package/dist/runtime/index.d.ts +1 -1
- package/dist/runtime/index.d.ts.map +1 -1
- package/dist/runtime/index.js.map +1 -1
- package/dist/runtime/manager.d.ts +128 -0
- package/dist/runtime/manager.d.ts.map +1 -1
- package/dist/runtime/manager.js +375 -64
- package/dist/runtime/manager.js.map +1 -1
- package/dist/runtime/types.d.ts +92 -1
- package/dist/runtime/types.d.ts.map +1 -1
- package/dist/search/evaluator.d.ts +71 -0
- package/dist/search/evaluator.d.ts.map +1 -1
- package/dist/search/evaluator.js +40 -5
- package/dist/search/evaluator.js.map +1 -1
- package/dist/search/index.d.ts +3 -5
- package/dist/search/index.d.ts.map +1 -1
- package/dist/search/index.js +4 -7
- package/dist/search/index.js.map +1 -1
- package/dist/search/moe-gate.d.ts +48 -2
- package/dist/search/moe-gate.d.ts.map +1 -1
- package/dist/search/moe-gate.js +117 -4
- package/dist/search/moe-gate.js.map +1 -1
- package/dist/search/refinement-loop.d.ts +63 -3
- package/dist/search/refinement-loop.d.ts.map +1 -1
- package/dist/search/refinement-loop.js +278 -23
- package/dist/search/refinement-loop.js.map +1 -1
- package/dist/search/refinement-types.d.ts +4 -2
- package/dist/search/refinement-types.d.ts.map +1 -1
- package/dist/search/refinement-types.js +2 -11
- package/dist/search/refinement-types.js.map +1 -1
- package/dist/search/router.d.ts +41 -34
- package/dist/search/router.d.ts.map +1 -1
- package/dist/search/router.js +156 -90
- package/dist/search/router.js.map +1 -1
- package/dist/search/solver.js +1 -1
- package/dist/search/solver.js.map +1 -1
- package/dist/search/task-difficulty.d.ts +54 -0
- package/dist/search/task-difficulty.d.ts.map +1 -0
- package/dist/search/task-difficulty.js +116 -0
- package/dist/search/task-difficulty.js.map +1 -0
- package/dist/search/tau-bench-verifier.d.ts +74 -0
- package/dist/search/tau-bench-verifier.d.ts.map +1 -0
- package/dist/search/tau-bench-verifier.js +96 -0
- package/dist/search/tau-bench-verifier.js.map +1 -0
- package/dist/search/verification-runner.d.ts.map +1 -1
- package/dist/search/verification-runner.js +6 -5
- package/dist/search/verification-runner.js.map +1 -1
- package/dist/session-bank/git-reader.d.ts.map +1 -1
- package/dist/session-bank/git-reader.js +5 -19
- package/dist/session-bank/git-reader.js.map +1 -1
- package/dist/session-bank/parser.d.ts.map +1 -1
- package/dist/session-bank/parser.js +26 -0
- package/dist/session-bank/parser.js.map +1 -1
- package/dist/session-bank/session-bank.d.ts.map +1 -1
- package/dist/session-bank/session-bank.js +2 -5
- package/dist/session-bank/session-bank.js.map +1 -1
- package/dist/session-bank/types.d.ts +24 -0
- package/dist/session-bank/types.d.ts.map +1 -1
- package/dist/surfacing/index.d.ts +3 -3
- package/dist/surfacing/index.d.ts.map +1 -1
- package/dist/surfacing/index.js +2 -2
- package/dist/surfacing/index.js.map +1 -1
- package/dist/surfacing/shared-bank.d.ts +43 -0
- package/dist/surfacing/shared-bank.d.ts.map +1 -0
- package/dist/surfacing/shared-bank.js +104 -0
- package/dist/surfacing/shared-bank.js.map +1 -0
- package/dist/surfacing/skill-library.d.ts +98 -2
- package/dist/surfacing/skill-library.d.ts.map +1 -1
- package/dist/surfacing/skill-library.js +192 -17
- package/dist/surfacing/skill-library.js.map +1 -1
- package/dist/surfacing/skill-publisher.d.ts +6 -4
- package/dist/surfacing/skill-publisher.d.ts.map +1 -1
- package/dist/surfacing/skill-publisher.js +32 -30
- package/dist/surfacing/skill-publisher.js.map +1 -1
- package/dist/surfacing/sqlite-storage-adapter.d.ts.map +1 -1
- package/dist/surfacing/sqlite-storage-adapter.js +12 -8
- package/dist/surfacing/sqlite-storage-adapter.js.map +1 -1
- package/dist/types/analysis.d.ts +43 -0
- package/dist/types/analysis.d.ts.map +1 -0
- package/dist/types/analysis.js +10 -0
- package/dist/types/analysis.js.map +1 -0
- package/dist/types/config.d.ts +2178 -425
- package/dist/types/config.d.ts.map +1 -1
- package/dist/types/config.js +600 -88
- package/dist/types/config.js.map +1 -1
- package/dist/types/dataclaw.d.ts +3 -3
- package/dist/types/index.d.ts +4 -7
- package/dist/types/index.d.ts.map +1 -1
- package/dist/types/index.js +2 -10
- package/dist/types/index.js.map +1 -1
- package/dist/types/knowledge-graph.d.ts +1 -1
- package/dist/types/knowledge-graph.d.ts.map +1 -1
- package/dist/types/knowledge.d.ts +219 -11
- package/dist/types/knowledge.d.ts.map +1 -1
- package/dist/types/knowledge.js +90 -30
- package/dist/types/knowledge.js.map +1 -1
- package/dist/types/memory.d.ts +8 -0
- package/dist/types/memory.d.ts.map +1 -1
- package/dist/types/memory.js +12 -4
- package/dist/types/memory.js.map +1 -1
- package/dist/types/outcome.d.ts +11 -0
- package/dist/types/outcome.d.ts.map +1 -1
- package/dist/types/outcome.js +9 -0
- package/dist/types/outcome.js.map +1 -1
- package/dist/types/playbook.d.ts +63 -0
- package/dist/types/playbook.d.ts.map +1 -1
- package/dist/types/playbook.js +51 -0
- package/dist/types/playbook.js.map +1 -1
- package/dist/types/trajectory-source.d.ts +1 -1
- package/dist/types/trajectory-source.d.ts.map +1 -1
- package/dist/types/trajectory.d.ts +61 -0
- package/dist/types/trajectory.d.ts.map +1 -1
- package/dist/types/trajectory.js +25 -0
- package/dist/types/trajectory.js.map +1 -1
- package/dist/utils/clock.d.ts +36 -0
- package/dist/utils/clock.d.ts.map +1 -0
- package/dist/utils/clock.js +48 -0
- package/dist/utils/clock.js.map +1 -0
- package/dist/utils/frontmatter.d.ts.map +1 -1
- package/dist/utils/frontmatter.js +10 -2
- package/dist/utils/frontmatter.js.map +1 -1
- package/dist/utils/index.d.ts +4 -1
- package/dist/utils/index.d.ts.map +1 -1
- package/dist/utils/index.js +4 -1
- package/dist/utils/index.js.map +1 -1
- package/dist/utils/logger.d.ts +31 -0
- package/dist/utils/logger.d.ts.map +1 -0
- package/dist/utils/logger.js +50 -0
- package/dist/utils/logger.js.map +1 -0
- package/dist/utils/objectives.d.ts +45 -0
- package/dist/utils/objectives.d.ts.map +1 -0
- package/dist/utils/objectives.js +74 -0
- package/dist/utils/objectives.js.map +1 -0
- package/dist/utils/partitioned-store.d.ts.map +1 -1
- package/dist/utils/partitioned-store.js.map +1 -1
- package/dist/utils/playbook-hash.d.ts +19 -0
- package/dist/utils/playbook-hash.d.ts.map +1 -0
- package/dist/utils/playbook-hash.js +51 -0
- package/dist/utils/playbook-hash.js.map +1 -0
- package/dist/utils/similarity.d.ts.map +1 -1
- package/dist/utils/similarity.js.map +1 -1
- package/dist/utils/solver-canonicalizer.d.ts +22 -0
- package/dist/utils/solver-canonicalizer.d.ts.map +1 -0
- package/dist/utils/solver-canonicalizer.js +41 -0
- package/dist/utils/solver-canonicalizer.js.map +1 -0
- package/dist/utils/storage.d.ts.map +1 -1
- package/dist/utils/storage.js +1 -3
- package/dist/utils/storage.js.map +1 -1
- package/dist/utils/task-signature.d.ts +34 -0
- package/dist/utils/task-signature.d.ts.map +1 -0
- package/dist/utils/task-signature.js +112 -0
- package/dist/utils/task-signature.js.map +1 -0
- package/dist/utils/trajectory-helpers.d.ts.map +1 -1
- package/dist/utils/trajectory-helpers.js.map +1 -1
- package/dist/utils/validation.d.ts.map +1 -1
- package/dist/utils/validation.js.map +1 -1
- package/dist/workspace/efficacy-toolkit.d.ts.map +1 -1
- package/dist/workspace/efficacy-toolkit.js +18 -26
- package/dist/workspace/efficacy-toolkit.js.map +1 -1
- package/dist/workspace/index.d.ts +4 -3
- package/dist/workspace/index.d.ts.map +1 -1
- package/dist/workspace/index.js +5 -3
- package/dist/workspace/index.js.map +1 -1
- package/dist/workspace/runner.d.ts.map +1 -1
- package/dist/workspace/runner.js +95 -21
- package/dist/workspace/runner.js.map +1 -1
- package/dist/workspace/skill-converter.d.ts +66 -1
- package/dist/workspace/skill-converter.d.ts.map +1 -1
- package/dist/workspace/skill-converter.js +238 -19
- package/dist/workspace/skill-converter.js.map +1 -1
- package/dist/workspace/skill-import.d.ts +43 -0
- package/dist/workspace/skill-import.d.ts.map +1 -0
- package/dist/workspace/skill-import.js +40 -0
- package/dist/workspace/skill-import.js.map +1 -0
- package/dist/workspace/templates/index.d.ts +6 -8
- package/dist/workspace/templates/index.d.ts.map +1 -1
- package/dist/workspace/templates/index.js +7 -11
- package/dist/workspace/templates/index.js.map +1 -1
- package/dist/workspace/templates/knowledge-defrag.d.ts.map +1 -1
- package/dist/workspace/templates/knowledge-defrag.js +0 -2
- package/dist/workspace/templates/knowledge-defrag.js.map +1 -1
- package/dist/workspace/templates/knowledge-extraction.d.ts +11 -0
- package/dist/workspace/templates/knowledge-extraction.d.ts.map +1 -1
- package/dist/workspace/templates/knowledge-extraction.js +11 -6
- package/dist/workspace/templates/knowledge-extraction.js.map +1 -1
- package/dist/workspace/templates/playbook-conditioning.d.ts +38 -0
- package/dist/workspace/templates/playbook-conditioning.d.ts.map +1 -0
- package/dist/workspace/templates/playbook-conditioning.js +147 -0
- package/dist/workspace/templates/playbook-conditioning.js.map +1 -0
- package/dist/workspace/templates/playbook-decay-detection.d.ts.map +1 -1
- package/dist/workspace/templates/playbook-decay-detection.js +24 -18
- package/dist/workspace/templates/playbook-decay-detection.js.map +1 -1
- package/dist/workspace/templates/playbook-efficacy-audit.d.ts.map +1 -1
- package/dist/workspace/templates/playbook-efficacy-audit.js +26 -11
- package/dist/workspace/templates/playbook-efficacy-audit.js.map +1 -1
- package/dist/workspace/templates/playbook-extraction.d.ts.map +1 -1
- package/dist/workspace/templates/playbook-extraction.js +75 -15
- package/dist/workspace/templates/playbook-extraction.js.map +1 -1
- package/dist/workspace/templates/playbook-lifecycle-review.d.ts.map +1 -1
- package/dist/workspace/templates/playbook-lifecycle-review.js +12 -8
- package/dist/workspace/templates/playbook-lifecycle-review.js.map +1 -1
- package/dist/workspace/templates/refinement-analysis.d.ts.map +1 -1
- package/dist/workspace/templates/refinement-analysis.js +11 -5
- package/dist/workspace/templates/refinement-analysis.js.map +1 -1
- package/dist/workspace/templates/skill-enrichment.d.ts.map +1 -1
- package/dist/workspace/templates/skill-enrichment.js +7 -11
- package/dist/workspace/templates/skill-enrichment.js.map +1 -1
- package/dist/workspace/templates/solution-evaluation.d.ts.map +1 -1
- package/dist/workspace/templates/solution-evaluation.js +15 -8
- package/dist/workspace/templates/solution-evaluation.js.map +1 -1
- package/dist/workspace/templates/trajectory-analysis.d.ts.map +1 -1
- package/dist/workspace/templates/trajectory-analysis.js +1 -3
- package/dist/workspace/templates/trajectory-analysis.js.map +1 -1
- package/dist/workspace/templates/usage-inference.d.ts.map +1 -1
- package/dist/workspace/templates/usage-inference.js +9 -7
- package/dist/workspace/templates/usage-inference.js.map +1 -1
- package/dist/workspace/types.d.ts +32 -2
- package/dist/workspace/types.d.ts.map +1 -1
- package/dist/workspace/types.js +29 -0
- package/dist/workspace/types.js.map +1 -1
- package/package.json +43 -19
- package/playbooks/agent-ops/long-task-checkpointing.json +54 -0
- package/playbooks/agent-ops/progressive-context-gathering.json +56 -0
- package/playbooks/agent-ops/safe-destructive-operations.json +59 -0
- package/playbooks/agent-ops/scope-discipline.json +56 -0
- package/playbooks/agent-ops/stall-detection-strategy-switch.json +55 -0
- package/playbooks/agent-ops/verification-before-completion.json +54 -0
- package/playbooks/alfworld/clean-and-place.json +35 -0
- package/playbooks/alfworld/cool-and-place.json +35 -0
- package/playbooks/alfworld/examine-in-light.json +32 -0
- package/playbooks/alfworld/heat-and-place.json +35 -0
- package/playbooks/alfworld/pick-and-place.json +37 -0
- package/playbooks/alfworld/pick-two-and-place.json +35 -0
- package/.claude/settings.json +0 -197
- package/.entire/settings.json +0 -4
- package/.sessionlog/settings.json +0 -4
- package/CLAUDE.md +0 -233
- package/SKILL.md +0 -193
- package/docs/DESIGN-agentic-workspace.md +0 -2057
- package/docs/DESIGN-semantic-memory-knowledge-bank.md +0 -1789
- package/docs/DESIGN-session-bank.md +0 -1134
- package/docs/DESIGN-team-extraction-pipeline.md +0 -1369
- package/docs/PROPOSAL-team-aware-learning.md +0 -1080
- package/docs/RESEARCH-semantic-memory-knowledge-structures.md +0 -517
- package/docs/RESEARCH-team-trajectory-learning.md +0 -553
- package/gaps.md +0 -204
- package/scripts/migrate-to-playbooks.ts +0 -307
- package/src/agents/index.ts +0 -14
- package/src/agents/mock-provider.ts +0 -93
- package/src/agents/types.ts +0 -137
- package/src/atlas.ts +0 -1198
- package/src/bin/cli-utils.ts +0 -220
- package/src/bin/cognitive-core.ts +0 -162
- package/src/bin/commands/kb.ts +0 -266
- package/src/bin/commands/learn.ts +0 -100
- package/src/bin/commands/legacy.ts +0 -182
- package/src/bin/commands/run.ts +0 -113
- package/src/bin/commands/sessions.ts +0 -221
- package/src/bin/commands/skills.ts +0 -146
- package/src/embeddings/bm25.ts +0 -337
- package/src/embeddings/index.ts +0 -42
- package/src/embeddings/inverted-index.ts +0 -134
- package/src/embeddings/manager.ts +0 -293
- package/src/embeddings/provider.ts +0 -318
- package/src/embeddings/vector-store.ts +0 -365
- package/src/factory.ts +0 -280
- package/src/index.ts +0 -495
- package/src/learning/analyzer.ts +0 -319
- package/src/learning/effectiveness.ts +0 -428
- package/src/learning/energy-evaluator.ts +0 -282
- package/src/learning/healing-orchestrator.ts +0 -383
- package/src/learning/index.ts +0 -150
- package/src/learning/instant-loop.ts +0 -357
- package/src/learning/knowledge-extractor.ts +0 -470
- package/src/learning/maintenance-scheduler.ts +0 -271
- package/src/learning/meta-learner.ts +0 -552
- package/src/learning/playbook-extractor.ts +0 -702
- package/src/learning/reflexion-generator.ts +0 -273
- package/src/learning/team-ingester.ts +0 -499
- package/src/learning/team-meta-learner.ts +0 -558
- package/src/learning/team-pipeline.ts +0 -364
- package/src/learning/trajectory-sources/dataclaw.ts +0 -403
- package/src/learning/trajectory-sources/entire.ts +0 -251
- package/src/learning/trajectory-sources/file.ts +0 -136
- package/src/learning/trajectory-sources/huggingface.ts +0 -248
- package/src/learning/trajectory-sources/in-memory.ts +0 -61
- package/src/learning/trajectory-sources/index.ts +0 -15
- package/src/learning/trajectory-sources/pipeline.ts +0 -69
- package/src/learning/unified-pipeline.ts +0 -1191
- package/src/learning/usage-inference.ts +0 -372
- package/src/mcp/index.ts +0 -12
- package/src/mcp/playbook-server.ts +0 -565
- package/src/memory/candidate-retrieval.ts +0 -72
- package/src/memory/causal-store.ts +0 -273
- package/src/memory/coherence.ts +0 -252
- package/src/memory/curated-loader.ts +0 -213
- package/src/memory/experience.ts +0 -682
- package/src/memory/graph-layers/base.ts +0 -184
- package/src/memory/graph-layers/causal.ts +0 -16
- package/src/memory/graph-layers/entity.ts +0 -16
- package/src/memory/graph-layers/index.ts +0 -5
- package/src/memory/graph-layers/semantic.ts +0 -16
- package/src/memory/graph-layers/temporal.ts +0 -16
- package/src/memory/index.ts +0 -131
- package/src/memory/knowledge-bank.ts +0 -1317
- package/src/memory/knowledge-defrag.ts +0 -329
- package/src/memory/knowledge-graph.ts +0 -361
- package/src/memory/meta.ts +0 -488
- package/src/memory/playbook.ts +0 -676
- package/src/memory/reasoning-bank.ts +0 -458
- package/src/memory/reflexion.ts +0 -122
- package/src/memory/search-provider.ts +0 -35
- package/src/memory/search-providers/index.ts +0 -3
- package/src/memory/search-providers/minimem.ts +0 -84
- package/src/memory/search-providers/text-similarity.ts +0 -35
- package/src/memory/skill-exporter.ts +0 -357
- package/src/memory/source-resolver.ts +0 -422
- package/src/memory/system.ts +0 -620
- package/src/memory/team-experience.ts +0 -604
- package/src/memory/temporal-compressor.ts +0 -409
- package/src/persistence/index.ts +0 -37
- package/src/persistence/migrator.ts +0 -298
- package/src/persistence/serializers.ts +0 -79
- package/src/persistence/sqlite-persistence.ts +0 -925
- package/src/runtime/backends/acp-protocol.ts +0 -231
- package/src/runtime/backends/acp.ts +0 -239
- package/src/runtime/backends/index.ts +0 -38
- package/src/runtime/backends/macro-agent.ts +0 -177
- package/src/runtime/backends/mock.ts +0 -218
- package/src/runtime/backends/subprocess.ts +0 -356
- package/src/runtime/compute-provider.ts +0 -206
- package/src/runtime/flows/learning.ts +0 -166
- package/src/runtime/flows/validation.ts +0 -381
- package/src/runtime/index.ts +0 -75
- package/src/runtime/manager.ts +0 -685
- package/src/runtime/types.ts +0 -198
- package/src/search/evaluator.ts +0 -499
- package/src/search/index.ts +0 -78
- package/src/search/moe-gate.ts +0 -304
- package/src/search/refinement-loop.ts +0 -457
- package/src/search/refinement-types.ts +0 -159
- package/src/search/router.ts +0 -368
- package/src/search/solver.ts +0 -303
- package/src/search/team-router.ts +0 -459
- package/src/search/verification-runner.ts +0 -570
- package/src/session-bank/git-reader.ts +0 -200
- package/src/session-bank/index.ts +0 -26
- package/src/session-bank/parser.ts +0 -548
- package/src/session-bank/session-bank.ts +0 -485
- package/src/session-bank/types.ts +0 -176
- package/src/surfacing/index.ts +0 -29
- package/src/surfacing/publisher.ts +0 -23
- package/src/surfacing/skill-library.ts +0 -625
- package/src/surfacing/skill-publisher.ts +0 -258
- package/src/surfacing/sqlite-storage-adapter.ts +0 -287
- package/src/surfacing/team-skill-library.ts +0 -610
- package/src/types/config.ts +0 -661
- package/src/types/dataclaw.ts +0 -99
- package/src/types/huggingface-transformers.d.ts +0 -12
- package/src/types/index.ts +0 -293
- package/src/types/knowledge-graph.ts +0 -246
- package/src/types/knowledge.ts +0 -388
- package/src/types/memory.ts +0 -300
- package/src/types/meta.ts +0 -218
- package/src/types/outcome.ts +0 -66
- package/src/types/playbook.ts +0 -215
- package/src/types/step.ts +0 -40
- package/src/types/task.ts +0 -52
- package/src/types/team-meta.ts +0 -212
- package/src/types/team-playbook.ts +0 -384
- package/src/types/team-trajectory.ts +0 -673
- package/src/types/trajectory-source.ts +0 -47
- package/src/types/trajectory.ts +0 -80
- package/src/utils/error-classifier.ts +0 -113
- package/src/utils/frontmatter.ts +0 -118
- package/src/utils/index.ts +0 -63
- package/src/utils/partitioned-store.ts +0 -299
- package/src/utils/similarity.ts +0 -139
- package/src/utils/storage.ts +0 -249
- package/src/utils/trajectory-helpers.ts +0 -79
- package/src/utils/validation.ts +0 -286
- package/src/workspace/efficacy-toolkit.ts +0 -496
- package/src/workspace/index.ts +0 -77
- package/src/workspace/runner.ts +0 -281
- package/src/workspace/skill-converter.ts +0 -288
- package/src/workspace/templates/index.ts +0 -95
- package/src/workspace/templates/knowledge-defrag.ts +0 -223
- package/src/workspace/templates/knowledge-extraction.ts +0 -337
- package/src/workspace/templates/meta-reflection.ts +0 -208
- package/src/workspace/templates/playbook-decay-detection.ts +0 -272
- package/src/workspace/templates/playbook-efficacy-audit.ts +0 -246
- package/src/workspace/templates/playbook-extraction.ts +0 -240
- package/src/workspace/templates/playbook-lifecycle-review.ts +0 -274
- package/src/workspace/templates/refinement-analysis.ts +0 -162
- package/src/workspace/templates/skill-enrichment.ts +0 -275
- package/src/workspace/templates/solution-evaluation.ts +0 -199
- package/src/workspace/templates/team-playbook-extraction.ts +0 -631
- package/src/workspace/templates/team-trajectory-analysis.ts +0 -564
- package/src/workspace/templates/trajectory-analysis.ts +0 -234
- package/src/workspace/templates/usage-inference.ts +0 -163
- package/src/workspace/types.ts +0 -185
- package/tests/atlas-knowledge.test.ts +0 -165
- package/tests/embeddings/bm25.test.ts +0 -130
- package/tests/embeddings/inverted-index.test.ts +0 -138
- package/tests/embeddings/manager.test.ts +0 -205
- package/tests/feature-toggles.test.ts +0 -275
- package/tests/fixtures/behavioral-trajectories.ts +0 -210
- package/tests/gap-fixes.test.ts +0 -1116
- package/tests/integration/atlas.test.ts +0 -266
- package/tests/integration/cli-e2e.test.ts +0 -621
- package/tests/integration/curated-sources-e2e.test.ts +0 -502
- package/tests/integration/dataclaw-e2e.test.ts +0 -559
- package/tests/integration/e2e.test.ts +0 -1337
- package/tests/integration/entire-e2e.test.ts +0 -376
- package/tests/integration/huggingface-e2e.test.ts +0 -627
- package/tests/integration/persistence-e2e.test.ts +0 -741
- package/tests/integration/phase-e2e.test.ts +0 -1143
- package/tests/integration/pipeline-data-correctness.test.ts +0 -794
- package/tests/integration/ranking-driven-loadout-e2e.test.ts +0 -185
- package/tests/integration/session-bank.test.ts +0 -231
- package/tests/integration/sessionlog-e2e.test.ts +0 -329
- package/tests/integration/skill-publishing-filesystem-e2e.test.ts +0 -216
- package/tests/integration/skill-tree-wiring.test.ts +0 -152
- package/tests/integration/unified-pipeline-e2e.test.ts +0 -634
- package/tests/learning/analyzer.test.ts +0 -426
- package/tests/learning/dataclaw-trajectory-source.test.ts +0 -642
- package/tests/learning/effectiveness.test.ts +0 -542
- package/tests/learning/energy-evaluator.test.ts +0 -180
- package/tests/learning/entire-trajectory-source.test.ts +0 -380
- package/tests/learning/healing-orchestrator.test.ts +0 -269
- package/tests/learning/huggingface-trajectory-source.test.ts +0 -817
- package/tests/learning/instant-loop.test.ts +0 -243
- package/tests/learning/knowledge-extractor.test.ts +0 -491
- package/tests/learning/maintenance-scheduler.test.ts +0 -191
- package/tests/learning/meta-learner.test.ts +0 -418
- package/tests/learning/pipeline-memory-updates.test.ts +0 -721
- package/tests/learning/playbook-extractor-provenance.test.ts +0 -114
- package/tests/learning/reflexion-generator.test.ts +0 -411
- package/tests/learning/team-ingester.test.ts +0 -349
- package/tests/learning/team-meta-learner.test.ts +0 -618
- package/tests/learning/team-pipeline.test.ts +0 -334
- package/tests/learning/trajectory-sources.test.ts +0 -320
- package/tests/learning/unified-pipeline-efficacy.test.ts +0 -232
- package/tests/learning/unified-pipeline.test.ts +0 -322
- package/tests/learning/usage-inference.test.ts +0 -254
- package/tests/mcp/playbook-server.test.ts +0 -257
- package/tests/memory/candidate-retrieval.test.ts +0 -167
- package/tests/memory/causal-store.test.ts +0 -276
- package/tests/memory/coherence.test.ts +0 -232
- package/tests/memory/compound-engineering-seed.test.ts +0 -338
- package/tests/memory/curated-loader-extended.test.ts +0 -225
- package/tests/memory/experience.test.ts +0 -203
- package/tests/memory/graph-layers/semantic.test.ts +0 -219
- package/tests/memory/knowledge-bank-extraction.test.ts +0 -558
- package/tests/memory/knowledge-bank.test.ts +0 -705
- package/tests/memory/knowledge-defrag.test.ts +0 -366
- package/tests/memory/knowledge-evolution.test.ts +0 -563
- package/tests/memory/knowledge-graph.test.ts +0 -492
- package/tests/memory/knowledge-inbox.test.ts +0 -258
- package/tests/memory/knowledge-minimem.test.ts +0 -251
- package/tests/memory/meta.test.ts +0 -399
- package/tests/memory/playbook-quality-validation.test.ts +0 -430
- package/tests/memory/playbook.test.ts +0 -644
- package/tests/memory/provenance.test.ts +0 -648
- package/tests/memory/reasoning-bank.test.ts +0 -239
- package/tests/memory/reflexion.test.ts +0 -166
- package/tests/memory/skill-exporter.test.ts +0 -475
- package/tests/memory/source-resolver.test.ts +0 -700
- package/tests/memory/system.test.ts +0 -330
- package/tests/memory/team-experience.test.ts +0 -411
- package/tests/memory/temporal-compressor.test.ts +0 -318
- package/tests/persistence/migrator.test.ts +0 -1009
- package/tests/persistence/sqlite-persistence.test.ts +0 -635
- package/tests/runtime/agent-manager.test.ts +0 -517
- package/tests/runtime/compute-provider.test.ts +0 -288
- package/tests/runtime/delegate.test.ts +0 -354
- package/tests/runtime/macro-agent-backend.test.ts +0 -266
- package/tests/runtime/mock-backend.test.ts +0 -248
- package/tests/search/evaluator.test.ts +0 -257
- package/tests/search/moe-gate.test.ts +0 -250
- package/tests/search/refinement-loop.test.ts +0 -477
- package/tests/search/refinement.test.ts +0 -267
- package/tests/search/router.test.ts +0 -506
- package/tests/search/team-router.test.ts +0 -376
- package/tests/search/verification-runner.test.ts +0 -357
- package/tests/session-bank/fixtures/multi-tool-session/full.jsonl +0 -12
- package/tests/session-bank/fixtures/multi-tool-session/metadata.json +0 -28
- package/tests/session-bank/fixtures/root-metadata.json +0 -18
- package/tests/session-bank/fixtures/sessionlog-root-metadata.json +0 -16
- package/tests/session-bank/fixtures/sessionlog-session/full.jsonl +0 -6
- package/tests/session-bank/fixtures/sessionlog-session/metadata.json +0 -55
- package/tests/session-bank/fixtures/simple-session/full.jsonl +0 -6
- package/tests/session-bank/fixtures/simple-session/metadata.json +0 -38
- package/tests/session-bank/git-reader.test.ts +0 -232
- package/tests/session-bank/parser.test.ts +0 -453
- package/tests/session-bank/session-bank.test.ts +0 -546
- package/tests/surfacing/skill-library.test.ts +0 -297
- package/tests/surfacing/skill-publisher.test.ts +0 -432
- package/tests/surfacing/sqlite-storage-adapter.test.ts +0 -197
- package/tests/surfacing/team-skill-library.test.ts +0 -444
- package/tests/types/outcome.test.ts +0 -147
- package/tests/types/step.test.ts +0 -133
- package/tests/types/task.test.ts +0 -158
- package/tests/types/team-meta.test.ts +0 -147
- package/tests/types/team-playbook.test.ts +0 -246
- package/tests/types/team-trajectory.test.ts +0 -557
- package/tests/types/trajectory.test.ts +0 -253
- package/tests/utils/error-classifier.test.ts +0 -149
- package/tests/utils/frontmatter.test.ts +0 -208
- package/tests/utils/partitioned-store.test.ts +0 -230
- package/tests/utils/similarity.test.ts +0 -188
- package/tests/utils/trajectory-helpers.test.ts +0 -163
- package/tests/utils/validation.test.ts +0 -252
- package/tests/workspace/efficacy-toolkit.test.ts +0 -404
- package/tests/workspace/full-flow.test.ts +0 -845
- package/tests/workspace/manager.test.ts +0 -215
- package/tests/workspace/runner.test.ts +0 -336
- package/tests/workspace/skill-converter.test.ts +0 -205
- package/tests/workspace/templates/knowledge-extraction.test.ts +0 -235
- package/tests/workspace/templates/playbook-efficacy.test.ts +0 -377
- package/tests/workspace/templates/team-playbook-extraction.test.ts +0 -341
- package/tests/workspace/templates/team-trajectory-analysis.test.ts +0 -417
- package/tsconfig.json +0 -25
- package/vitest.config.ts +0 -22
|
@@ -0,0 +1,169 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Per-family difficulty registry + discriminative task sampler (Phase B(2) of
|
|
3
|
+
* the adaptive eval-economics roadmap — DESIGN-meta-learning-validation.md §10).
|
|
4
|
+
*
|
|
5
|
+
* Motivation: episodes spent on tasks the model always solves (or never
|
|
6
|
+
* solves) carry almost no information about a config delta — Fisher
|
|
7
|
+
* information for a binary outcome peaks at p(success) ≈ 0.5. The registry
|
|
8
|
+
* accumulates per-family success counts across runs; the sampler then weights
|
|
9
|
+
* task selection toward the discriminative band and away from
|
|
10
|
+
* saturated-either-way families.
|
|
11
|
+
*
|
|
12
|
+
* The family unit is `skillTag`: it is stamped on every TaskResult row (so
|
|
13
|
+
* historical artifacts are ingestible) and it is the unit that repeats across
|
|
14
|
+
* runs even when benchmark instances rotate. This is what makes the registry
|
|
15
|
+
* work "in the wild": a never-seen instance inherits its family's prior
|
|
16
|
+
* (hierarchical pooling on features), rather than requiring per-instance
|
|
17
|
+
* repetition.
|
|
18
|
+
*
|
|
19
|
+
* Difficulty is estimated from cold-arm counts when enough exist (learning
|
|
20
|
+
* shifts p over a run, contaminating the estimate), falling back to pooled
|
|
21
|
+
* counts when the cold sample is thin.
|
|
22
|
+
*
|
|
23
|
+
* Priors are ADVISORY: a family with fewer than `minRuns` ingested runs gets
|
|
24
|
+
* weight 1 (uniform), and known families keep a weight floor rather than
|
|
25
|
+
* being hard-dropped — per the design doc's small-n IRT risk note.
|
|
26
|
+
*
|
|
27
|
+
* Estimand caveat: discriminative selection changes what a success rate
|
|
28
|
+
* means. Numbers from a discriminative stream are comparable to each other,
|
|
29
|
+
* not to fixed-stream runs — always report which sampler produced a number.
|
|
30
|
+
*/
|
|
31
|
+
import { existsSync, mkdirSync, readFileSync, renameSync, writeFileSync } from 'node:fs';
|
|
32
|
+
import { dirname } from 'node:path';
|
|
33
|
+
import { seededRng } from '../rng.js';
|
|
34
|
+
const COLD_ARM_LABEL = 'no-learning';
|
|
35
|
+
/** Runs per family before the discriminative weight becomes non-uniform. */
|
|
36
|
+
const DEFAULT_MIN_RUNS = 3;
|
|
37
|
+
/** Cold-arm episodes needed before cold counts alone drive the estimate. */
|
|
38
|
+
const MIN_COLD_ATTEMPTS = 4;
|
|
39
|
+
/** Weight floor for known families — advisory down-weighting, never a drop. */
|
|
40
|
+
const WEIGHT_FLOOR = 0.05;
|
|
41
|
+
export function emptyDifficultyRegistry() {
|
|
42
|
+
return { version: 1, updatedAt: new Date(0).toISOString(), benchmarks: {} };
|
|
43
|
+
}
|
|
44
|
+
/** Load a registry file, returning an empty registry when the file is absent. */
|
|
45
|
+
export function loadDifficultyRegistry(path) {
|
|
46
|
+
if (!existsSync(path))
|
|
47
|
+
return emptyDifficultyRegistry();
|
|
48
|
+
const raw = JSON.parse(readFileSync(path, 'utf8'));
|
|
49
|
+
if (raw.version !== 1) {
|
|
50
|
+
throw new Error(`Unsupported difficulty registry version ${String(raw.version)} in ${path}`);
|
|
51
|
+
}
|
|
52
|
+
return raw;
|
|
53
|
+
}
|
|
54
|
+
/** Atomic write (temp + rename), matching the checkpoint writer's discipline. */
|
|
55
|
+
export function saveDifficultyRegistry(path, registry) {
|
|
56
|
+
registry.updatedAt = new Date().toISOString();
|
|
57
|
+
mkdirSync(dirname(path), { recursive: true });
|
|
58
|
+
const tmp = `${path}.tmp`;
|
|
59
|
+
writeFileSync(tmp, JSON.stringify(registry, null, 2));
|
|
60
|
+
renameSync(tmp, path);
|
|
61
|
+
}
|
|
62
|
+
/**
|
|
63
|
+
* Fold one run's episode rows into the registry. `runs` is incremented once
|
|
64
|
+
* per family per call, so a call corresponds to one run artifact.
|
|
65
|
+
*/
|
|
66
|
+
export function ingestRunRows(registry, benchmark, rows) {
|
|
67
|
+
const bench = (registry.benchmarks[benchmark] ??= { families: {} });
|
|
68
|
+
const touched = new Set();
|
|
69
|
+
for (const row of rows) {
|
|
70
|
+
if (row.status === 'env_error')
|
|
71
|
+
continue;
|
|
72
|
+
const stats = (bench.families[row.family] ??= {
|
|
73
|
+
attempts: 0,
|
|
74
|
+
successes: 0,
|
|
75
|
+
coldAttempts: 0,
|
|
76
|
+
coldSuccesses: 0,
|
|
77
|
+
runs: 0,
|
|
78
|
+
});
|
|
79
|
+
stats.attempts++;
|
|
80
|
+
if (row.success)
|
|
81
|
+
stats.successes++;
|
|
82
|
+
if (row.arm === COLD_ARM_LABEL) {
|
|
83
|
+
stats.coldAttempts++;
|
|
84
|
+
if (row.success)
|
|
85
|
+
stats.coldSuccesses++;
|
|
86
|
+
}
|
|
87
|
+
touched.add(row.family);
|
|
88
|
+
}
|
|
89
|
+
for (const family of touched) {
|
|
90
|
+
bench.families[family].runs++;
|
|
91
|
+
}
|
|
92
|
+
}
|
|
93
|
+
/**
|
|
94
|
+
* Extract ingestible rows from a run artifact: either a final result JSON
|
|
95
|
+
* (learning.perSeed + noLearning.perSeed) or a <out>.checkpoint.json partial.
|
|
96
|
+
* Returns the benchmark name found in the artifact alongside the rows.
|
|
97
|
+
*/
|
|
98
|
+
export function extractIngestRows(raw) {
|
|
99
|
+
const rows = [];
|
|
100
|
+
if (raw.partial === true) {
|
|
101
|
+
const file = raw;
|
|
102
|
+
for (const t of file.tasks ?? []) {
|
|
103
|
+
rows.push({ family: t.skillTag, arm: t.arm, success: t.success, status: t.status });
|
|
104
|
+
}
|
|
105
|
+
const meta = (raw.meta ?? {});
|
|
106
|
+
return { benchmark: typeof meta.benchmark === 'string' ? meta.benchmark : undefined, rows };
|
|
107
|
+
}
|
|
108
|
+
const pullArm = (arm, condition) => {
|
|
109
|
+
const perSeed = condition?.perSeed ?? [];
|
|
110
|
+
for (const seedRows of perSeed) {
|
|
111
|
+
for (const r of seedRows) {
|
|
112
|
+
rows.push({ family: r.skillTag, arm, success: r.success, status: r.status });
|
|
113
|
+
}
|
|
114
|
+
}
|
|
115
|
+
};
|
|
116
|
+
pullArm('learning', raw.learning);
|
|
117
|
+
pullArm(COLD_ARM_LABEL, raw.noLearning);
|
|
118
|
+
return {
|
|
119
|
+
benchmark: typeof raw.benchmarkName === 'string' ? raw.benchmarkName : undefined,
|
|
120
|
+
rows,
|
|
121
|
+
};
|
|
122
|
+
}
|
|
123
|
+
/**
|
|
124
|
+
* Posterior-mean success estimate for a family under a Beta(1,1) prior.
|
|
125
|
+
* Cold-arm counts drive the estimate once they are thick enough; otherwise
|
|
126
|
+
* pooled counts (learning contamination beats no data).
|
|
127
|
+
*/
|
|
128
|
+
export function estimateFamilySuccess(stats) {
|
|
129
|
+
const useCold = stats.coldAttempts >= MIN_COLD_ATTEMPTS;
|
|
130
|
+
const attempts = useCold ? stats.coldAttempts : stats.attempts;
|
|
131
|
+
const successes = useCold ? stats.coldSuccesses : stats.successes;
|
|
132
|
+
return (successes + 1) / (attempts + 2);
|
|
133
|
+
}
|
|
134
|
+
/**
|
|
135
|
+
* Sampling weight for a family: 4·p̂(1−p̂), which is 1 at p̂=0.5 and falls
|
|
136
|
+
* toward 0 as the family saturates in either direction. Unknown or
|
|
137
|
+
* under-observed families (< minRuns ingested runs) get weight 1 — the
|
|
138
|
+
* registry is advisory until it has seen enough evidence.
|
|
139
|
+
*/
|
|
140
|
+
export function discriminativeWeight(stats, minRuns = DEFAULT_MIN_RUNS) {
|
|
141
|
+
if (!stats || stats.runs < minRuns)
|
|
142
|
+
return 1;
|
|
143
|
+
const p = estimateFamilySuccess(stats);
|
|
144
|
+
return Math.max(WEIGHT_FLOOR, 4 * p * (1 - p));
|
|
145
|
+
}
|
|
146
|
+
/**
|
|
147
|
+
* Weighted sampling without replacement (Efraimidis–Spirakis: key = u^(1/w),
|
|
148
|
+
* keep the `count` largest keys), deterministic per (registry, benchmark,
|
|
149
|
+
* seed). Selected tasks keep their original relative order — ordered
|
|
150
|
+
* benchmarks stay ordered, just with gaps — and are re-indexed 0..n-1 so
|
|
151
|
+
* positional aggregation (curves, checkpoint rows) stays consistent.
|
|
152
|
+
*/
|
|
153
|
+
export function selectDiscriminativeTasks(pool, count, registry, benchmark, seed, minRuns = DEFAULT_MIN_RUNS) {
|
|
154
|
+
if (pool.length <= count)
|
|
155
|
+
return pool;
|
|
156
|
+
const families = registry.benchmarks[benchmark]?.families ?? {};
|
|
157
|
+
const rng = seededRng('difficulty-select', benchmark, seed);
|
|
158
|
+
const keyed = pool.map((task, poolIdx) => {
|
|
159
|
+
const w = discriminativeWeight(families[task.skillTag], minRuns);
|
|
160
|
+
const u = Math.max(rng(), Number.MIN_VALUE);
|
|
161
|
+
return { poolIdx, key: Math.pow(u, 1 / w) };
|
|
162
|
+
});
|
|
163
|
+
keyed.sort((a, b) => b.key - a.key);
|
|
164
|
+
const chosen = new Set(keyed.slice(0, count).map((k) => k.poolIdx));
|
|
165
|
+
return pool
|
|
166
|
+
.filter((_, poolIdx) => chosen.has(poolIdx))
|
|
167
|
+
.map((task, index) => ({ ...task, index }));
|
|
168
|
+
}
|
|
169
|
+
//# sourceMappingURL=difficulty-registry.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"difficulty-registry.js","sourceRoot":"","sources":["../../../src/eval/live/difficulty-registry.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AACH,OAAO,EAAE,UAAU,EAAE,SAAS,EAAE,YAAY,EAAE,UAAU,EAAE,aAAa,EAAE,MAAM,SAAS,CAAC;AACzF,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AACpC,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AA+BtC,MAAM,cAAc,GAAG,aAAa,CAAC;AACrC,4EAA4E;AAC5E,MAAM,gBAAgB,GAAG,CAAC,CAAC;AAC3B,4EAA4E;AAC5E,MAAM,iBAAiB,GAAG,CAAC,CAAC;AAC5B,+EAA+E;AAC/E,MAAM,YAAY,GAAG,IAAI,CAAC;AAE1B,MAAM,UAAU,uBAAuB;IACrC,OAAO,EAAE,OAAO,EAAE,CAAC,EAAE,SAAS,EAAE,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,WAAW,EAAE,EAAE,UAAU,EAAE,EAAE,EAAE,CAAC;AAC9E,CAAC;AAED,iFAAiF;AACjF,MAAM,UAAU,sBAAsB,CAAC,IAAY;IACjD,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC;QAAE,OAAO,uBAAuB,EAAE,CAAC;IACxD,MAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,YAAY,CAAC,IAAI,EAAE,MAAM,CAAC,CAA2B,CAAC;IAC7E,IAAI,GAAG,CAAC,OAAO,KAAK,CAAC,EAAE,CAAC;QACtB,MAAM,IAAI,KAAK,CAAC,2CAA2C,MAAM,CAAC,GAAG,CAAC,OAAO,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC;IAC/F,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,iFAAiF;AACjF,MAAM,UAAU,sBAAsB,CAAC,IAAY,EAAE,QAAgC;IACnF,QAAQ,CAAC,SAAS,GAAG,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;IAC9C,SAAS,CAAC,OAAO,CAAC,IAAI,CAAC,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAC9C,MAAM,GAAG,GAAG,GAAG,IAAI,MAAM,CAAC;IAC1B,aAAa,CAAC,GAAG,EAAE,IAAI,CAAC,SAAS,CAAC,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;IACtD,UAAU,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC;AACxB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,aAAa,CAC3B,QAAgC,EAChC,SAAiB,EACjB,IAAiB;IAEjB,MAAM,KAAK,GAAG,CAAC,QAAQ,CAAC,UAAU,CAAC,SAAS,CAAC,KAAK,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC,CAAC;IACpE,MAAM,OAAO,GAAG,IAAI,GAAG,EAAU,CAAC;IAClC,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QACvB,IAAI,GAAG,CAAC,MAAM,KAAK,WAAW;YAAE,SAAS;QACzC,MAAM,KAAK,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,GAAG,CAAC,MAAM,CAAC,KAAK;YAC5C,QAAQ,EAAE,CAAC;YACX,SAAS,EAAE,CAAC;YACZ,YAAY,EAAE,CAAC;YACf,aAAa,EAAE,CAAC;YAChB,IAAI,EAAE,CAAC;SACR,CAAC,CAAC;QACH,KAAK,CAAC,QAAQ,EAAE,CAAC;QACjB,IAAI,GAAG,CAAC,OAAO;YAAE,KAAK,CAAC,SAAS,EAAE,CAAC;QACnC,IAAI,GAAG,CAAC,GAAG,KAAK,cAAc,EAAE,CAAC;YAC/B,KAAK,CAAC,YAAY,EAAE,CAAC;YACrB,IAAI,GAAG,CAAC,OAAO;gBAAE,KAAK,CAAC,aAAa,EAAE,CAAC;QACzC,CAAC;QACD,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;IAC1B,CAAC;IACD,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;QAC7B,KAAK,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,CAAC;IAChC,CAAC;AACH,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,iBAAiB,CAAC,GAA4B;IAI5D,MAAM,IAAI,GAAgB,EAAE,CAAC;IAE7B,IAAI,GAAG,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;QACzB,MAAM,IAAI,GAAG,GAAgC,CAAC;QAC9C,KAAK,MAAM,CAAC,IAAI,IAAI,CAAC,KAAK,IAAI,EAAE,EAAE,CAAC;YACjC,IAAI,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,CAAC,GAAG,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,EAAE,MAAM,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;QACtF,CAAC;QACD,MAAM,IAAI,GAAG,CAAC,GAAG,CAAC,IAAI,IAAI,EAAE,CAA4B,CAAC;QACzD,OAAO,EAAE,SAAS,EAAE,OAAO,IAAI,CAAC,SAAS,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,CAAC,SAAS,EAAE,IAAI,EAAE,CAAC;IAC9F,CAAC;IAED,MAAM,OAAO,GAAG,CAAC,GAAW,EAAE,SAAkB,EAAQ,EAAE;QACxD,MAAM,OAAO,GAAI,SAAsD,EAAE,OAAO,IAAI,EAAE,CAAC;QACvF,KAAK,MAAM,QAAQ,IAAI,OAAO,EAAE,CAAC;YAC/B,KAAK,MAAM,CAAC,IAAI,QAAQ,EAAE,CAAC;gBACzB,IAAI,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,QAAQ,EAAE,GAAG,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,EAAE,MAAM,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;YAC/E,CAAC;QACH,CAAC;IACH,CAAC,CAAC;IACF,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC,QAAQ,CAAC,CAAC;IAClC,OAAO,CAAC,cAAc,EAAE,GAAG,CAAC,UAAU,CAAC,CAAC;IACxC,OAAO;QACL,SAAS,EAAE,OAAO,GAAG,CAAC,aAAa,KAAK,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC,aAAa,CAAC,CAAC,CAAC,SAAS;QAChF,IAAI;KACL,CAAC;AACJ,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,qBAAqB,CAAC,KAA4B;IAChE,MAAM,OAAO,GAAG,KAAK,CAAC,YAAY,IAAI,iBAAiB,CAAC;IACxD,MAAM,QAAQ,GAAG,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC,KAAK,CAAC,QAAQ,CAAC;IAC/D,MAAM,SAAS,GAAG,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,aAAa,CAAC,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC;IAClE,OAAO,CAAC,SAAS,GAAG,CAAC,CAAC,GAAG,CAAC,QAAQ,GAAG,CAAC,CAAC,CAAC;AAC1C,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,oBAAoB,CAClC,KAAwC,EACxC,OAAO,GAAG,gBAAgB;IAE1B,IAAI,CAAC,KAAK,IAAI,KAAK,CAAC,IAAI,GAAG,OAAO;QAAE,OAAO,CAAC,CAAC;IAC7C,MAAM,CAAC,GAAG,qBAAqB,CAAC,KAAK,CAAC,CAAC;IACvC,OAAO,IAAI,CAAC,GAAG,CAAC,YAAY,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;AACjD,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,yBAAyB,CACvC,IAAgB,EAChB,KAAa,EACb,QAAgC,EAChC,SAAiB,EACjB,IAAY,EACZ,OAAO,GAAG,gBAAgB;IAE1B,IAAI,IAAI,CAAC,MAAM,IAAI,KAAK;QAAE,OAAO,IAAI,CAAC;IACtC,MAAM,QAAQ,GAAG,QAAQ,CAAC,UAAU,CAAC,SAAS,CAAC,EAAE,QAAQ,IAAI,EAAE,CAAC;IAChE,MAAM,GAAG,GAAG,SAAS,CAAC,mBAAmB,EAAE,SAAS,EAAE,IAAI,CAAC,CAAC;IAE5D,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,OAAO,EAAE,EAAE;QACvC,MAAM,CAAC,GAAG,oBAAoB,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,OAAO,CAAC,CAAC;QACjE,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,EAAE,MAAM,CAAC,SAAS,CAAC,CAAC;QAC5C,OAAO,EAAE,OAAO,EAAE,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC;IAC9C,CAAC,CAAC,CAAC;IACH,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC;IACpC,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC;IAEpE,OAAO,IAAI;SACR,MAAM,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,EAAE,CAAC,MAAM,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;SAC3C,GAAG,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC,EAAE,GAAG,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC;AAChD,CAAC"}
|
|
@@ -0,0 +1,52 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Docker orchestration for benchmark evaluation.
|
|
3
|
+
*
|
|
4
|
+
* Shells out to the `docker` CLI (same subprocess pattern as claude-cli.ts).
|
|
5
|
+
* Each benchmark task gets its own container to prevent state leakage.
|
|
6
|
+
*/
|
|
7
|
+
export interface DockerTaskConfig {
|
|
8
|
+
image: string;
|
|
9
|
+
mounts?: Array<{
|
|
10
|
+
hostPath: string;
|
|
11
|
+
containerPath: string;
|
|
12
|
+
}>;
|
|
13
|
+
env?: Record<string, string>;
|
|
14
|
+
workdir?: string;
|
|
15
|
+
timeoutMs?: number;
|
|
16
|
+
}
|
|
17
|
+
export interface DockerRunResult {
|
|
18
|
+
exitCode: number;
|
|
19
|
+
stdout: string;
|
|
20
|
+
stderr: string;
|
|
21
|
+
durationMs: number;
|
|
22
|
+
}
|
|
23
|
+
export declare class DockerOrchestrator {
|
|
24
|
+
private readonly opts;
|
|
25
|
+
private readonly builtImages;
|
|
26
|
+
constructor(opts?: {
|
|
27
|
+
pullPolicy?: 'always' | 'if-not-present';
|
|
28
|
+
});
|
|
29
|
+
buildImage(dockerfilePath: string, tag: string, context?: string): Promise<string>;
|
|
30
|
+
run(config: DockerTaskConfig, command: string[]): Promise<DockerRunResult>;
|
|
31
|
+
copyToContainer(containerId: string, hostPath: string, containerPath: string): Promise<void>;
|
|
32
|
+
runTests(config: DockerTaskConfig, testCommand: string[]): Promise<{
|
|
33
|
+
passed: boolean;
|
|
34
|
+
output: string;
|
|
35
|
+
}>;
|
|
36
|
+
cleanup(tag: string): Promise<void>;
|
|
37
|
+
startPersistentContainer(config: DockerTaskConfig & {
|
|
38
|
+
name: string;
|
|
39
|
+
}): Promise<string>;
|
|
40
|
+
execInContainer(name: string, command: string, timeoutMs?: number): Promise<DockerRunResult>;
|
|
41
|
+
copyInto(name: string, hostPath: string, containerPath: string): Promise<void>;
|
|
42
|
+
stopAndRemove(name: string): Promise<void>;
|
|
43
|
+
private exec;
|
|
44
|
+
private execWithResult;
|
|
45
|
+
private execOnce;
|
|
46
|
+
}
|
|
47
|
+
export declare function createDockerOrchestrator(opts?: {
|
|
48
|
+
pullPolicy?: 'always' | 'if-not-present';
|
|
49
|
+
}): DockerOrchestrator;
|
|
50
|
+
/** Quick check whether the `docker` binary is available. */
|
|
51
|
+
export declare function dockerAvailable(): Promise<boolean>;
|
|
52
|
+
//# sourceMappingURL=docker.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"docker.d.ts","sourceRoot":"","sources":["../../../src/eval/live/docker.ts"],"names":[],"mappings":"AAEA;;;;;GAKG;AAEH,MAAM,WAAW,gBAAgB;IAC/B,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,KAAK,CAAC;QAAE,QAAQ,EAAE,MAAM,CAAC;QAAC,aAAa,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAC5D,GAAG,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC7B,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,eAAe;IAC9B,QAAQ,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,MAAM,CAAC;IACf,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;CACpB;AAED,qBAAa,kBAAkB;IAGjB,OAAO,CAAC,QAAQ,CAAC,IAAI;IAFjC,OAAO,CAAC,QAAQ,CAAC,WAAW,CAAqB;gBAEpB,IAAI,GAAE;QAAE,UAAU,CAAC,EAAE,QAAQ,GAAG,gBAAgB,CAAA;KAAO;IAE9E,UAAU,CAAC,cAAc,EAAE,MAAM,EAAE,GAAG,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC;IAUlF,GAAG,CAAC,MAAM,EAAE,gBAAgB,EAAE,OAAO,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,eAAe,CAAC;IAiB1E,eAAe,CACnB,WAAW,EAAE,MAAM,EACnB,QAAQ,EAAE,MAAM,EAChB,aAAa,EAAE,MAAM,GACpB,OAAO,CAAC,IAAI,CAAC;IAIV,QAAQ,CACZ,MAAM,EAAE,gBAAgB,EACxB,WAAW,EAAE,MAAM,EAAE,GACpB,OAAO,CAAC;QAAE,MAAM,EAAE,OAAO,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IAQzC,OAAO,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC;IASnC,wBAAwB,CAAC,MAAM,EAAE,gBAAgB,GAAG;QAAE,IAAI,EAAE,MAAM,CAAA;KAAE,GAAG,OAAO,CAAC,MAAM,CAAC;IAqBtF,eAAe,CACnB,IAAI,EAAE,MAAM,EACZ,OAAO,EAAE,MAAM,EACf,SAAS,SAAU,GAClB,OAAO,CAAC,eAAe,CAAC;IAIrB,QAAQ,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,EAAE,aAAa,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC;IAO9E,aAAa,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC;IAahD,OAAO,CAAC,IAAI;YAIE,cAAc;IAmB5B,OAAO,CAAC,QAAQ;CA0CjB;AAED,wBAAgB,wBAAwB,CAAC,IAAI,CAAC,EAAE;IAC9C,UAAU,CAAC,EAAE,QAAQ,GAAG,gBAAgB,CAAC;CAC1C,GAAG,kBAAkB,CAErB;AAED,4DAA4D;AAC5D,wBAAsB,eAAe,IAAI,OAAO,CAAC,OAAO,CAAC,CAMxD"}
|
|
@@ -0,0 +1,164 @@
|
|
|
1
|
+
import { spawn } from 'node:child_process';
|
|
2
|
+
export class DockerOrchestrator {
|
|
3
|
+
opts;
|
|
4
|
+
builtImages = new Set();
|
|
5
|
+
constructor(opts = {}) {
|
|
6
|
+
this.opts = opts;
|
|
7
|
+
}
|
|
8
|
+
async buildImage(dockerfilePath, tag, context) {
|
|
9
|
+
if (this.opts.pullPolicy === 'if-not-present' && this.builtImages.has(tag)) {
|
|
10
|
+
return tag;
|
|
11
|
+
}
|
|
12
|
+
const args = ['build', '-t', tag, '-f', dockerfilePath, context ?? '.'];
|
|
13
|
+
await this.exec(args, 600_000);
|
|
14
|
+
this.builtImages.add(tag);
|
|
15
|
+
return tag;
|
|
16
|
+
}
|
|
17
|
+
async run(config, command) {
|
|
18
|
+
const args = ['run', '--rm'];
|
|
19
|
+
if (config.workdir)
|
|
20
|
+
args.push('-w', config.workdir);
|
|
21
|
+
if (config.mounts) {
|
|
22
|
+
for (const m of config.mounts) {
|
|
23
|
+
args.push('-v', `${m.hostPath}:${m.containerPath}`);
|
|
24
|
+
}
|
|
25
|
+
}
|
|
26
|
+
if (config.env) {
|
|
27
|
+
for (const [k, v] of Object.entries(config.env)) {
|
|
28
|
+
args.push('-e', `${k}=${v}`);
|
|
29
|
+
}
|
|
30
|
+
}
|
|
31
|
+
args.push(config.image, ...command);
|
|
32
|
+
return this.execWithResult(args, config.timeoutMs ?? 300_000);
|
|
33
|
+
}
|
|
34
|
+
async copyToContainer(containerId, hostPath, containerPath) {
|
|
35
|
+
await this.exec(['cp', hostPath, `${containerId}:${containerPath}`], 30_000);
|
|
36
|
+
}
|
|
37
|
+
async runTests(config, testCommand) {
|
|
38
|
+
const result = await this.run(config, testCommand);
|
|
39
|
+
return {
|
|
40
|
+
passed: result.exitCode === 0,
|
|
41
|
+
output: result.stdout + (result.stderr ? '\n' + result.stderr : ''),
|
|
42
|
+
};
|
|
43
|
+
}
|
|
44
|
+
async cleanup(tag) {
|
|
45
|
+
try {
|
|
46
|
+
await this.exec(['rmi', '-f', tag], 30_000);
|
|
47
|
+
}
|
|
48
|
+
catch {
|
|
49
|
+
// best-effort cleanup
|
|
50
|
+
}
|
|
51
|
+
this.builtImages.delete(tag);
|
|
52
|
+
}
|
|
53
|
+
async startPersistentContainer(config) {
|
|
54
|
+
const args = ['run', '-d', '--name', config.name];
|
|
55
|
+
if (config.workdir)
|
|
56
|
+
args.push('-w', config.workdir);
|
|
57
|
+
if (config.mounts) {
|
|
58
|
+
for (const m of config.mounts) {
|
|
59
|
+
args.push('-v', `${m.hostPath}:${m.containerPath}`);
|
|
60
|
+
}
|
|
61
|
+
}
|
|
62
|
+
if (config.env) {
|
|
63
|
+
for (const [k, v] of Object.entries(config.env)) {
|
|
64
|
+
args.push('-e', `${k}=${v}`);
|
|
65
|
+
}
|
|
66
|
+
}
|
|
67
|
+
args.push(config.image, 'tail', '-f', '/dev/null');
|
|
68
|
+
const result = await this.execOnce(args, config.timeoutMs ?? 60_000);
|
|
69
|
+
if (result.exitCode !== 0) {
|
|
70
|
+
throw new Error(`Failed to start container ${config.name}: ${result.stderr}`);
|
|
71
|
+
}
|
|
72
|
+
return config.name;
|
|
73
|
+
}
|
|
74
|
+
async execInContainer(name, command, timeoutMs = 120_000) {
|
|
75
|
+
return this.execOnce(['exec', name, 'bash', '-c', command], timeoutMs);
|
|
76
|
+
}
|
|
77
|
+
async copyInto(name, hostPath, containerPath) {
|
|
78
|
+
const result = await this.execOnce(['cp', hostPath, `${name}:${containerPath}`], 60_000);
|
|
79
|
+
if (result.exitCode !== 0) {
|
|
80
|
+
throw new Error(`docker cp failed: ${result.stderr}`);
|
|
81
|
+
}
|
|
82
|
+
}
|
|
83
|
+
async stopAndRemove(name) {
|
|
84
|
+
try {
|
|
85
|
+
await this.execOnce(['stop', '-t', '5', name], 15_000);
|
|
86
|
+
}
|
|
87
|
+
catch {
|
|
88
|
+
/* best effort */
|
|
89
|
+
}
|
|
90
|
+
try {
|
|
91
|
+
await this.execOnce(['rm', '-f', name], 15_000);
|
|
92
|
+
}
|
|
93
|
+
catch {
|
|
94
|
+
/* best effort */
|
|
95
|
+
}
|
|
96
|
+
}
|
|
97
|
+
exec(args, timeoutMs, retries = 2) {
|
|
98
|
+
return this.execWithResult(args, timeoutMs, retries).then(() => { });
|
|
99
|
+
}
|
|
100
|
+
async execWithResult(args, timeoutMs, retries = 2) {
|
|
101
|
+
let lastErr;
|
|
102
|
+
for (let attempt = 0; attempt <= retries; attempt++) {
|
|
103
|
+
try {
|
|
104
|
+
return await this.execOnce(args, timeoutMs);
|
|
105
|
+
}
|
|
106
|
+
catch (err) {
|
|
107
|
+
lastErr = err;
|
|
108
|
+
if (attempt < retries) {
|
|
109
|
+
await new Promise((r) => setTimeout(r, 3000 * (attempt + 1)));
|
|
110
|
+
}
|
|
111
|
+
}
|
|
112
|
+
}
|
|
113
|
+
throw lastErr ?? new Error('docker command failed');
|
|
114
|
+
}
|
|
115
|
+
execOnce(args, timeoutMs) {
|
|
116
|
+
return new Promise((resolve, reject) => {
|
|
117
|
+
const t0 = Date.now();
|
|
118
|
+
const child = spawn('docker', args, { stdio: ['pipe', 'pipe', 'pipe'] });
|
|
119
|
+
let stdout = '';
|
|
120
|
+
let stderr = '';
|
|
121
|
+
let settled = false;
|
|
122
|
+
const timer = setTimeout(() => {
|
|
123
|
+
if (settled)
|
|
124
|
+
return;
|
|
125
|
+
settled = true;
|
|
126
|
+
child.kill('SIGKILL');
|
|
127
|
+
reject(new Error(`docker command timed out after ${timeoutMs}ms: docker ${args.slice(0, 3).join(' ')}...`));
|
|
128
|
+
}, timeoutMs);
|
|
129
|
+
child.stdout.on('data', (d) => (stdout += d.toString()));
|
|
130
|
+
child.stderr.on('data', (d) => (stderr += d.toString()));
|
|
131
|
+
child.on('error', (err) => {
|
|
132
|
+
if (settled)
|
|
133
|
+
return;
|
|
134
|
+
settled = true;
|
|
135
|
+
clearTimeout(timer);
|
|
136
|
+
reject(new Error(`Failed to spawn docker: ${err.message}`));
|
|
137
|
+
});
|
|
138
|
+
child.on('close', (code) => {
|
|
139
|
+
if (settled)
|
|
140
|
+
return;
|
|
141
|
+
settled = true;
|
|
142
|
+
clearTimeout(timer);
|
|
143
|
+
resolve({
|
|
144
|
+
exitCode: code ?? 1,
|
|
145
|
+
stdout,
|
|
146
|
+
stderr,
|
|
147
|
+
durationMs: Date.now() - t0,
|
|
148
|
+
});
|
|
149
|
+
});
|
|
150
|
+
});
|
|
151
|
+
}
|
|
152
|
+
}
|
|
153
|
+
export function createDockerOrchestrator(opts) {
|
|
154
|
+
return new DockerOrchestrator(opts);
|
|
155
|
+
}
|
|
156
|
+
/** Quick check whether the `docker` binary is available. */
|
|
157
|
+
export async function dockerAvailable() {
|
|
158
|
+
return new Promise((resolve) => {
|
|
159
|
+
const child = spawn('docker', ['version'], { stdio: 'ignore' });
|
|
160
|
+
child.on('error', () => resolve(false));
|
|
161
|
+
child.on('close', (code) => resolve(code === 0));
|
|
162
|
+
});
|
|
163
|
+
}
|
|
164
|
+
//# sourceMappingURL=docker.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"docker.js","sourceRoot":"","sources":["../../../src/eval/live/docker.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAC;AAwB3C,MAAM,OAAO,kBAAkB;IAGA;IAFZ,WAAW,GAAG,IAAI,GAAG,EAAU,CAAC;IAEjD,YAA6B,OAAqD,EAAE;QAAvD,SAAI,GAAJ,IAAI,CAAmD;IAAG,CAAC;IAExF,KAAK,CAAC,UAAU,CAAC,cAAsB,EAAE,GAAW,EAAE,OAAgB;QACpE,IAAI,IAAI,CAAC,IAAI,CAAC,UAAU,KAAK,gBAAgB,IAAI,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC;YAC3E,OAAO,GAAG,CAAC;QACb,CAAC;QACD,MAAM,IAAI,GAAG,CAAC,OAAO,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI,EAAE,cAAc,EAAE,OAAO,IAAI,GAAG,CAAC,CAAC;QACxE,MAAM,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QAC/B,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QAC1B,OAAO,GAAG,CAAC;IACb,CAAC;IAED,KAAK,CAAC,GAAG,CAAC,MAAwB,EAAE,OAAiB;QACnD,MAAM,IAAI,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC;QAC7B,IAAI,MAAM,CAAC,OAAO;YAAE,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,MAAM,CAAC,OAAO,CAAC,CAAC;QACpD,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;YAClB,KAAK,MAAM,CAAC,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;gBAC9B,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,QAAQ,IAAI,CAAC,CAAC,aAAa,EAAE,CAAC,CAAC;YACtD,CAAC;QACH,CAAC;QACD,IAAI,MAAM,CAAC,GAAG,EAAE,CAAC;YACf,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC;gBAChD,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YAC/B,CAAC;QACH,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,KAAK,EAAE,GAAG,OAAO,CAAC,CAAC;QACpC,OAAO,IAAI,CAAC,cAAc,CAAC,IAAI,EAAE,MAAM,CAAC,SAAS,IAAI,OAAO,CAAC,CAAC;IAChE,CAAC;IAED,KAAK,CAAC,eAAe,CACnB,WAAmB,EACnB,QAAgB,EAChB,aAAqB;QAErB,MAAM,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,QAAQ,EAAE,GAAG,WAAW,IAAI,aAAa,EAAE,CAAC,EAAE,MAAM,CAAC,CAAC;IAC/E,CAAC;IAED,KAAK,CAAC,QAAQ,CACZ,MAAwB,EACxB,WAAqB;QAErB,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;QACnD,OAAO;YACL,MAAM,EAAE,MAAM,CAAC,QAAQ,KAAK,CAAC;YAC7B,MAAM,EAAE,MAAM,CAAC,MAAM,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;SACpE,CAAC;IACJ,CAAC;IAED,KAAK,CAAC,OAAO,CAAC,GAAW;QACvB,IAAI,CAAC;YACH,MAAM,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,IAAI,EAAE,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC;QAC9C,CAAC;QAAC,MAAM,CAAC;YACP,sBAAsB;QACxB,CAAC;QACD,IAAI,CAAC,WAAW,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC;IAC/B,CAAC;IAED,KAAK,CAAC,wBAAwB,CAAC,MAA2C;QACxE,MAAM,IAAI,GAAG,CAAC,KAAK,EAAE,IAAI,EAAE,QAAQ,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC;QAClD,IAAI,MAAM,CAAC,OAAO;YAAE,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,MAAM,CAAC,OAAO,CAAC,CAAC;QACpD,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;YAClB,KAAK,MAAM,CAAC,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;gBAC9B,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,QAAQ,IAAI,CAAC,CAAC,aAAa,EAAE,CAAC,CAAC;YACtD,CAAC;QACH,CAAC;QACD,IAAI,MAAM,CAAC,GAAG,EAAE,CAAC;YACf,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC;gBAChD,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YAC/B,CAAC;QACH,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,WAAW,CAAC,CAAC;QACnD,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC,SAAS,IAAI,MAAM,CAAC,CAAC;QACrE,IAAI,MAAM,CAAC,QAAQ,KAAK,CAAC,EAAE,CAAC;YAC1B,MAAM,IAAI,KAAK,CAAC,6BAA6B,MAAM,CAAC,IAAI,KAAK,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;QAChF,CAAC;QACD,OAAO,MAAM,CAAC,IAAI,CAAC;IACrB,CAAC;IAED,KAAK,CAAC,eAAe,CACnB,IAAY,EACZ,OAAe,EACf,SAAS,GAAG,OAAO;QAEnB,OAAO,IAAI,CAAC,QAAQ,CAAC,CAAC,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,OAAO,CAAC,EAAE,SAAS,CAAC,CAAC;IACzE,CAAC;IAED,KAAK,CAAC,QAAQ,CAAC,IAAY,EAAE,QAAgB,EAAE,aAAqB;QAClE,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,QAAQ,CAAC,CAAC,IAAI,EAAE,QAAQ,EAAE,GAAG,IAAI,IAAI,aAAa,EAAE,CAAC,EAAE,MAAM,CAAC,CAAC;QACzF,IAAI,MAAM,CAAC,QAAQ,KAAK,CAAC,EAAE,CAAC;YAC1B,MAAM,IAAI,KAAK,CAAC,qBAAqB,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;QACxD,CAAC;IACH,CAAC;IAED,KAAK,CAAC,aAAa,CAAC,IAAY;QAC9B,IAAI,CAAC;YACH,MAAM,IAAI,CAAC,QAAQ,CAAC,CAAC,MAAM,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI,CAAC,EAAE,MAAM,CAAC,CAAC;QACzD,CAAC;QAAC,MAAM,CAAC;YACP,iBAAiB;QACnB,CAAC;QACD,IAAI,CAAC;YACH,MAAM,IAAI,CAAC,QAAQ,CAAC,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC,EAAE,MAAM,CAAC,CAAC;QAClD,CAAC;QAAC,MAAM,CAAC;YACP,iBAAiB;QACnB,CAAC;IACH,CAAC;IAEO,IAAI,CAAC,IAAc,EAAE,SAAiB,EAAE,OAAO,GAAG,CAAC;QACzD,OAAO,IAAI,CAAC,cAAc,CAAC,IAAI,EAAE,SAAS,EAAE,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,GAAE,CAAC,CAAC,CAAC;IACtE,CAAC;IAEO,KAAK,CAAC,cAAc,CAC1B,IAAc,EACd,SAAiB,EACjB,OAAO,GAAG,CAAC;QAEX,IAAI,OAA0B,CAAC;QAC/B,KAAK,IAAI,OAAO,GAAG,CAAC,EAAE,OAAO,IAAI,OAAO,EAAE,OAAO,EAAE,EAAE,CAAC;YACpD,IAAI,CAAC;gBACH,OAAO,MAAM,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;YAC9C,CAAC;YAAC,OAAO,GAAG,EAAE,CAAC;gBACb,OAAO,GAAG,GAAY,CAAC;gBACvB,IAAI,OAAO,GAAG,OAAO,EAAE,CAAC;oBACtB,MAAM,IAAI,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,CAAC,EAAE,IAAI,GAAG,CAAC,OAAO,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;gBAChE,CAAC;YACH,CAAC;QACH,CAAC;QACD,MAAM,OAAO,IAAI,IAAI,KAAK,CAAC,uBAAuB,CAAC,CAAC;IACtD,CAAC;IAEO,QAAQ,CAAC,IAAc,EAAE,SAAiB;QAChD,OAAO,IAAI,OAAO,CAAkB,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;YACtD,MAAM,EAAE,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;YACtB,MAAM,KAAK,GAAG,KAAK,CAAC,QAAQ,EAAE,IAAI,EAAE,EAAE,KAAK,EAAE,CAAC,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC,CAAC;YACzE,IAAI,MAAM,GAAG,EAAE,CAAC;YAChB,IAAI,MAAM,GAAG,EAAE,CAAC;YAChB,IAAI,OAAO,GAAG,KAAK,CAAC;YAEpB,MAAM,KAAK,GAAG,UAAU,CAAC,GAAG,EAAE;gBAC5B,IAAI,OAAO;oBAAE,OAAO;gBACpB,OAAO,GAAG,IAAI,CAAC;gBACf,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;gBACtB,MAAM,CACJ,IAAI,KAAK,CACP,kCAAkC,SAAS,cAAc,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CACzF,CACF,CAAC;YACJ,CAAC,EAAE,SAAS,CAAC,CAAC;YAEd,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC;YACzD,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC;YAEzD,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE;gBACxB,IAAI,OAAO;oBAAE,OAAO;gBACpB,OAAO,GAAG,IAAI,CAAC;gBACf,YAAY,CAAC,KAAK,CAAC,CAAC;gBACpB,MAAM,CAAC,IAAI,KAAK,CAAC,2BAA2B,GAAG,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;YAC9D,CAAC,CAAC,CAAC;YAEH,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE;gBACzB,IAAI,OAAO;oBAAE,OAAO;gBACpB,OAAO,GAAG,IAAI,CAAC;gBACf,YAAY,CAAC,KAAK,CAAC,CAAC;gBACpB,OAAO,CAAC;oBACN,QAAQ,EAAE,IAAI,IAAI,CAAC;oBACnB,MAAM;oBACN,MAAM;oBACN,UAAU,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,EAAE;iBAC5B,CAAC,CAAC;YACL,CAAC,CAAC,CAAC;QACL,CAAC,CAAC,CAAC;IACL,CAAC;CACF;AAED,MAAM,UAAU,wBAAwB,CAAC,IAExC;IACC,OAAO,IAAI,kBAAkB,CAAC,IAAI,CAAC,CAAC;AACtC,CAAC;AAED,4DAA4D;AAC5D,MAAM,CAAC,KAAK,UAAU,eAAe;IACnC,OAAO,IAAI,OAAO,CAAC,CAAC,OAAO,EAAE,EAAE;QAC7B,MAAM,KAAK,GAAG,KAAK,CAAC,QAAQ,EAAE,CAAC,SAAS,CAAC,EAAE,EAAE,KAAK,EAAE,QAAQ,EAAE,CAAC,CAAC;QAChE,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,GAAG,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC;QACxC,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC;IACnD,CAAC,CAAC,CAAC;AACL,CAAC"}
|
|
@@ -0,0 +1,68 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* JSONL bridge to an external (Python) text-environment process, e.g. ALFWorld.
|
|
3
|
+
*
|
|
4
|
+
* Protocol (one JSON object per line, request → response):
|
|
5
|
+
* → {"cmd":"list"} ← {"tasks":["...", ...]}
|
|
6
|
+
* → {"cmd":"reset","task":"<id>"} ← {"obs":"...", "admissible":[...], "done":false}
|
|
7
|
+
* → {"cmd":"step","action":"go to ..."} ← {"obs":"...", "reward":0|1, "done":bool, "admissible":[...]}
|
|
8
|
+
*
|
|
9
|
+
* The bridge process is supplied via the ALFWORLD_BRIDGE env var (a command
|
|
10
|
+
* line) or defaults to `python <repo>/src/eval/live/bridges/alfworld_bridge.py`.
|
|
11
|
+
* This keeps cognitive-core free of any Python/ALFWorld dependency — the bridge
|
|
12
|
+
* is opt-in and only needed for the ALFWorld live arm.
|
|
13
|
+
*/
|
|
14
|
+
export declare class EnvBridge {
|
|
15
|
+
private readonly command;
|
|
16
|
+
private readonly args;
|
|
17
|
+
private readonly requestTimeoutMs;
|
|
18
|
+
/** Extra env vars merged over process.env for the spawned bridge process.
|
|
19
|
+
* Used by the SkillFlow e2b runtime to inject a computed sandbox timeout. */
|
|
20
|
+
private readonly env?;
|
|
21
|
+
private proc;
|
|
22
|
+
private rl;
|
|
23
|
+
private queue;
|
|
24
|
+
private dead;
|
|
25
|
+
constructor(command: string, args: string[], requestTimeoutMs?: number,
|
|
26
|
+
/** Extra env vars merged over process.env for the spawned bridge process.
|
|
27
|
+
* Used by the SkillFlow e2b runtime to inject a computed sandbox timeout. */
|
|
28
|
+
env?: Record<string, string> | undefined);
|
|
29
|
+
/** Start the bridge process. Waits for a ping handshake to confirm readiness. */
|
|
30
|
+
start(): Promise<void>;
|
|
31
|
+
protected request<T>(payload: Record<string, unknown>): Promise<T>;
|
|
32
|
+
listTasks(): Promise<{
|
|
33
|
+
tasks: string[];
|
|
34
|
+
}>;
|
|
35
|
+
reset(task: string): Promise<EnvObservation>;
|
|
36
|
+
step(action: string): Promise<EnvObservation>;
|
|
37
|
+
/**
|
|
38
|
+
* Send an arbitrary command to the bridge. Subclasses (e.g. SkillsBench,
|
|
39
|
+
* SkillFlow) use this to extend the JSONL protocol with benchmark-specific
|
|
40
|
+
* commands without duplicating the request plumbing.
|
|
41
|
+
*/
|
|
42
|
+
send<T>(payload: Record<string, unknown>): Promise<T>;
|
|
43
|
+
close(): Promise<void>;
|
|
44
|
+
}
|
|
45
|
+
/**
|
|
46
|
+
* Thrown when the Python bridge reports an env-side crash via the `error` field.
|
|
47
|
+
* Distinguishes infra flakiness from genuine agent failures so the runner can
|
|
48
|
+
* mark the task as `env_error` instead of `failure`.
|
|
49
|
+
*/
|
|
50
|
+
export declare class EnvBridgeError extends Error {
|
|
51
|
+
readonly isEnvBridgeError = true;
|
|
52
|
+
constructor(message: string);
|
|
53
|
+
}
|
|
54
|
+
export interface EnvObservation {
|
|
55
|
+
obs: string;
|
|
56
|
+
reward?: number;
|
|
57
|
+
done?: boolean;
|
|
58
|
+
admissible?: string[];
|
|
59
|
+
/** Set by the bridge when the env raised — distinguishes infra error from a
|
|
60
|
+
* genuine task failure (the bridge still emits done:true so the loop ends). */
|
|
61
|
+
error?: string;
|
|
62
|
+
}
|
|
63
|
+
/** Resolve the bridge command from ALFWORLD_BRIDGE or a default python invocation. */
|
|
64
|
+
export declare function resolveBridgeCommand(defaultScriptPath: string): {
|
|
65
|
+
command: string;
|
|
66
|
+
args: string[];
|
|
67
|
+
};
|
|
68
|
+
//# sourceMappingURL=env-bridge.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"env-bridge.d.ts","sourceRoot":"","sources":["../../../src/eval/live/env-bridge.ts"],"names":[],"mappings":"AAIA;;;;;;;;;;;;GAYG;AACH,qBAAa,SAAS;IAOlB,OAAO,CAAC,QAAQ,CAAC,OAAO;IACxB,OAAO,CAAC,QAAQ,CAAC,IAAI;IACrB,OAAO,CAAC,QAAQ,CAAC,gBAAgB;IACjC;iFAC6E;IAC7E,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC;IAXvB,OAAO,CAAC,IAAI,CAA8D;IAC1E,OAAO,CAAC,EAAE,CAA0B;IACpC,OAAO,CAAC,KAAK,CAAgF;IAC7F,OAAO,CAAC,IAAI,CAAS;gBAGF,OAAO,EAAE,MAAM,EACf,IAAI,EAAE,MAAM,EAAE,EACd,gBAAgB,SAAU;IAC3C;iFAC6E;IAC5D,GAAG,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,YAAA;IAG/C,iFAAiF;IAC3E,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;IA8B5B,SAAS,CAAC,OAAO,CAAC,CAAC,EAAE,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC;IA+ClE,SAAS,IAAI,OAAO,CAAC;QAAE,KAAK,EAAE,MAAM,EAAE,CAAA;KAAE,CAAC;IAIzC,KAAK,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,cAAc,CAAC;IAI5C,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAAC,cAAc,CAAC;IAI7C;;;;OAIG;IACH,IAAI,CAAC,CAAC,EAAE,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC;IAI/C,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;CAoC7B;AAED;;;;GAIG;AACH,qBAAa,cAAe,SAAQ,KAAK;IACvC,QAAQ,CAAC,gBAAgB,QAAQ;gBACrB,OAAO,EAAE,MAAM;CAI5B;AAED,MAAM,WAAW,cAAc;IAC7B,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,OAAO,CAAC;IACf,UAAU,CAAC,EAAE,MAAM,EAAE,CAAC;IACtB;mFAC+E;IAC/E,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;AAED,sFAAsF;AACtF,wBAAgB,oBAAoB,CAAC,iBAAiB,EAAE,MAAM,GAAG;IAC/D,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,EAAE,MAAM,EAAE,CAAC;CAChB,CAOA"}
|