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 +1,147 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Phase-2 memory provider for the skillopt-v1 head-to-head — the lever that
|
|
3
|
+
* turns ccore's GEPA-like proposer from "vanilla GEPA" into "GEPA + ccore memory".
|
|
4
|
+
*
|
|
5
|
+
* Lever 1 (retrieval-conditioned proposals), the dominant one: every optimization
|
|
6
|
+
* rollout is recorded into a real ccore `ExperienceMemory`; before each reflection
|
|
7
|
+
* the proposer is handed the top-k most-similar *successful* prior answers, so it
|
|
8
|
+
* can reuse a structure that already passed the checks instead of re-inventing it.
|
|
9
|
+
*
|
|
10
|
+
* Shared across seeds (lever 4, cross-seed accumulation) when the runner reuses one
|
|
11
|
+
* instance; within a single seed it already gives the two-timescale benefit (lever
|
|
12
|
+
* 3) — later iterations retrieve earlier iterations' successes. Held-out (`test`)
|
|
13
|
+
* rollouts are NEVER recorded (no leakage); the runner only records optimization
|
|
14
|
+
* rollouts.
|
|
15
|
+
*
|
|
16
|
+
* See docs/DESIGN-skillopt-headtohead.md §4.
|
|
17
|
+
*/
|
|
18
|
+
import { tmpdir } from 'node:os';
|
|
19
|
+
import { join } from 'node:path';
|
|
20
|
+
import { createSqlitePersistence } from '../../persistence/index.js';
|
|
21
|
+
import { ExperienceMemory } from '../../memory/experience.js';
|
|
22
|
+
/**
|
|
23
|
+
* Extract the structural skeleton of an answer: markdown headings, bold labels,
|
|
24
|
+
* and `Label:` lines — the elements section_present / regex checks look for.
|
|
25
|
+
*/
|
|
26
|
+
export function structuralSignature(answer) {
|
|
27
|
+
const sigs = new Set();
|
|
28
|
+
for (const raw of (answer || '').split('\n')) {
|
|
29
|
+
const heading = raw.match(/^\s{0,3}(#{1,6})\s+(.+?)\s*$/);
|
|
30
|
+
if (heading)
|
|
31
|
+
sigs.add(`${heading[1]} ${heading[2]}`.slice(0, 60));
|
|
32
|
+
const bold = raw.match(/^\s*\*\*(.+?)\*\*/);
|
|
33
|
+
if (bold)
|
|
34
|
+
sigs.add(`**${bold[1]}**`.slice(0, 60));
|
|
35
|
+
const label = raw.match(/^\s*([A-Z][A-Za-z ]{1,28}):(?:\s|$)/);
|
|
36
|
+
if (label)
|
|
37
|
+
sigs.add(`${label[1]}:`.slice(0, 60));
|
|
38
|
+
}
|
|
39
|
+
return [...sigs];
|
|
40
|
+
}
|
|
41
|
+
export class SkillOptV1Memory {
|
|
42
|
+
memory;
|
|
43
|
+
k;
|
|
44
|
+
domain;
|
|
45
|
+
successOnly;
|
|
46
|
+
mode;
|
|
47
|
+
maxExemplars;
|
|
48
|
+
excerptChars;
|
|
49
|
+
counter = 0;
|
|
50
|
+
constructor(persistence, cfg) {
|
|
51
|
+
this.memory = new ExperienceMemory(persistence);
|
|
52
|
+
this.k = cfg.k ?? 4;
|
|
53
|
+
this.domain = cfg.domain ?? 'skillopt-v1';
|
|
54
|
+
this.successOnly = cfg.successOnly ?? true;
|
|
55
|
+
this.mode = cfg.mode ?? 'recipe';
|
|
56
|
+
this.maxExemplars = cfg.maxExemplars ?? 6;
|
|
57
|
+
this.excerptChars = cfg.excerptChars ?? 240;
|
|
58
|
+
}
|
|
59
|
+
/** Build an isolated, ephemeral memory (own temp SQLite dir). */
|
|
60
|
+
static async create(cfg = {}) {
|
|
61
|
+
const baseDir = join(tmpdir(), `skillopt_v1_mem_${process.pid}_${Date.now()}`);
|
|
62
|
+
const persistence = createSqlitePersistence({ baseDir });
|
|
63
|
+
await persistence.init();
|
|
64
|
+
const m = new SkillOptV1Memory(persistence, cfg);
|
|
65
|
+
await m.memory.init();
|
|
66
|
+
return m;
|
|
67
|
+
}
|
|
68
|
+
/** Record one optimization rollout as an experience (success keyed on hard score). */
|
|
69
|
+
async record(task, output, score, feedback) {
|
|
70
|
+
const id = `exp-${this.domain}-${this.counter++}`;
|
|
71
|
+
const exp = {
|
|
72
|
+
id,
|
|
73
|
+
taskInput: task.intent,
|
|
74
|
+
solutionOutput: output,
|
|
75
|
+
feedback,
|
|
76
|
+
success: score >= 1,
|
|
77
|
+
domain: this.domain,
|
|
78
|
+
trajectoryId: id,
|
|
79
|
+
usageCount: 0,
|
|
80
|
+
accessScore: 0,
|
|
81
|
+
retrievalCount: 0,
|
|
82
|
+
createdAt: new Date(),
|
|
83
|
+
tier: 'hot',
|
|
84
|
+
metadata: { seed: task.id.split(':')[0] ?? '', split: task.split },
|
|
85
|
+
};
|
|
86
|
+
await this.memory.add(exp);
|
|
87
|
+
}
|
|
88
|
+
/** Record a whole evaluated batch (parallel arrays). */
|
|
89
|
+
async recordBatch(tasks, ev) {
|
|
90
|
+
for (let i = 0; i < tasks.length; i++) {
|
|
91
|
+
await this.record(tasks[i], ev.outputs[i] ?? '', ev.scores[i] ?? 0, ev.feedback[i] ?? '');
|
|
92
|
+
}
|
|
93
|
+
}
|
|
94
|
+
/**
|
|
95
|
+
* Retrieve top-k similar successful exemplars for a minibatch and render them
|
|
96
|
+
* as reflection context. Deduplicates by experience id across the batch.
|
|
97
|
+
*/
|
|
98
|
+
async context(tasks) {
|
|
99
|
+
// Gather distinct successful exemplars across the minibatch, capped.
|
|
100
|
+
const seen = new Set();
|
|
101
|
+
const exemplars = [];
|
|
102
|
+
for (const t of tasks) {
|
|
103
|
+
const hits = await this.memory.findSimilar(t.intent, {
|
|
104
|
+
k: this.k,
|
|
105
|
+
domain: this.domain,
|
|
106
|
+
successOnly: this.successOnly,
|
|
107
|
+
});
|
|
108
|
+
for (const { experience: e } of hits) {
|
|
109
|
+
if (seen.has(e.id))
|
|
110
|
+
continue;
|
|
111
|
+
seen.add(e.id);
|
|
112
|
+
exemplars.push({ taskInput: e.taskInput, output: e.solutionOutput });
|
|
113
|
+
if (exemplars.length >= this.maxExemplars)
|
|
114
|
+
break;
|
|
115
|
+
}
|
|
116
|
+
if (exemplars.length >= this.maxExemplars)
|
|
117
|
+
break;
|
|
118
|
+
}
|
|
119
|
+
if (exemplars.length === 0)
|
|
120
|
+
return '';
|
|
121
|
+
if (this.mode === 'recipe') {
|
|
122
|
+
// Compact: the union of structural elements the winners shared.
|
|
123
|
+
const sigs = new Set();
|
|
124
|
+
for (const e of exemplars)
|
|
125
|
+
for (const s of structuralSignature(e.output))
|
|
126
|
+
sigs.add(s);
|
|
127
|
+
if (sigs.size > 0) {
|
|
128
|
+
return [
|
|
129
|
+
'Prior answers that PASSED all checks on similar tasks consistently included',
|
|
130
|
+
'these structural elements — reproduce every one of them:',
|
|
131
|
+
...[...sigs].map((s) => `- ${s}`),
|
|
132
|
+
].join('\n');
|
|
133
|
+
}
|
|
134
|
+
// no extractable structure (e.g. max_chars/contains winners) → fall through
|
|
135
|
+
}
|
|
136
|
+
// excerpt mode (or recipe fallback): short raw excerpts.
|
|
137
|
+
return [
|
|
138
|
+
'These prior answers PASSED all automated checks on similar tasks. Reuse the',
|
|
139
|
+
'structure/sections/format that worked rather than re-inventing it:',
|
|
140
|
+
...exemplars.map((e) => `- Task: ${e.taskInput}\n Winning answer (excerpt): ${e.output.slice(0, this.excerptChars)}`),
|
|
141
|
+
].join('\n');
|
|
142
|
+
}
|
|
143
|
+
async close() {
|
|
144
|
+
await this.memory.close();
|
|
145
|
+
}
|
|
146
|
+
}
|
|
147
|
+
//# sourceMappingURL=skillopt_v1_memory.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"skillopt_v1_memory.js","sourceRoot":"","sources":["../../../src/eval/live/skillopt_v1_memory.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AACH,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AACjC,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AACjC,OAAO,EAAE,uBAAuB,EAA0B,MAAM,4BAA4B,CAAC;AAC7F,OAAO,EAAE,gBAAgB,EAAE,MAAM,4BAA4B,CAAC;AA4B9D;;;GAGG;AACH,MAAM,UAAU,mBAAmB,CAAC,MAAc;IAChD,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAC/B,KAAK,MAAM,GAAG,IAAI,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC;QAC7C,MAAM,OAAO,GAAG,GAAG,CAAC,KAAK,CAAC,8BAA8B,CAAC,CAAC;QAC1D,IAAI,OAAO;YAAE,IAAI,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC,IAAI,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;QAClE,MAAM,IAAI,GAAG,GAAG,CAAC,KAAK,CAAC,mBAAmB,CAAC,CAAC;QAC5C,IAAI,IAAI;YAAE,IAAI,CAAC,GAAG,CAAC,KAAK,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;QAClD,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC,qCAAqC,CAAC,CAAC;QAC/D,IAAI,KAAK;YAAE,IAAI,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;IACnD,CAAC;IACD,OAAO,CAAC,GAAG,IAAI,CAAC,CAAC;AACnB,CAAC;AAED,MAAM,OAAO,gBAAgB;IACV,MAAM,CAAmB;IACzB,CAAC,CAAS;IACV,MAAM,CAAS;IACf,WAAW,CAAU;IACrB,IAAI,CAAuB;IAC3B,YAAY,CAAS;IACrB,YAAY,CAAS;IAC9B,OAAO,GAAG,CAAC,CAAC;IAEpB,YAAoB,WAA8B,EAAE,GAA2B;QAC7E,IAAI,CAAC,MAAM,GAAG,IAAI,gBAAgB,CAAC,WAAW,CAAC,CAAC;QAChD,IAAI,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC;QACpB,IAAI,CAAC,MAAM,GAAG,GAAG,CAAC,MAAM,IAAI,aAAa,CAAC;QAC1C,IAAI,CAAC,WAAW,GAAG,GAAG,CAAC,WAAW,IAAI,IAAI,CAAC;QAC3C,IAAI,CAAC,IAAI,GAAG,GAAG,CAAC,IAAI,IAAI,QAAQ,CAAC;QACjC,IAAI,CAAC,YAAY,GAAG,GAAG,CAAC,YAAY,IAAI,CAAC,CAAC;QAC1C,IAAI,CAAC,YAAY,GAAG,GAAG,CAAC,YAAY,IAAI,GAAG,CAAC;IAC9C,CAAC;IAED,iEAAiE;IACjE,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,MAA8B,EAAE;QAClD,MAAM,OAAO,GAAG,IAAI,CAAC,MAAM,EAAE,EAAE,mBAAmB,OAAO,CAAC,GAAG,IAAI,IAAI,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;QAC/E,MAAM,WAAW,GAAG,uBAAuB,CAAC,EAAE,OAAO,EAAE,CAAC,CAAC;QACzD,MAAM,WAAW,CAAC,IAAI,EAAE,CAAC;QACzB,MAAM,CAAC,GAAG,IAAI,gBAAgB,CAAC,WAAW,EAAE,GAAG,CAAC,CAAC;QACjD,MAAM,CAAC,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;QACtB,OAAO,CAAC,CAAC;IACX,CAAC;IAED,sFAAsF;IACtF,KAAK,CAAC,MAAM,CACV,IAAoB,EACpB,MAAc,EACd,KAAa,EACb,QAAgB;QAEhB,MAAM,EAAE,GAAG,OAAO,IAAI,CAAC,MAAM,IAAI,IAAI,CAAC,OAAO,EAAE,EAAE,CAAC;QAClD,MAAM,GAAG,GAAe;YACtB,EAAE;YACF,SAAS,EAAE,IAAI,CAAC,MAAM;YACtB,cAAc,EAAE,MAAM;YACtB,QAAQ;YACR,OAAO,EAAE,KAAK,IAAI,CAAC;YACnB,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,YAAY,EAAE,EAAE;YAChB,UAAU,EAAE,CAAC;YACb,WAAW,EAAE,CAAC;YACd,cAAc,EAAE,CAAC;YACjB,SAAS,EAAE,IAAI,IAAI,EAAE;YACrB,IAAI,EAAE,KAAK;YACX,QAAQ,EAAE,EAAE,IAAI,EAAE,IAAI,CAAC,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,KAAK,EAAE,IAAI,CAAC,KAAK,EAAE;SACnE,CAAC;QACF,MAAM,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IAC7B,CAAC;IAED,wDAAwD;IACxD,KAAK,CAAC,WAAW,CAAC,KAAuB,EAAE,EAAkB;QAC3D,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACtC,MAAM,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;QAC5F,CAAC;IACH,CAAC;IAED;;;OAGG;IACH,KAAK,CAAC,OAAO,CAAC,KAAuB;QACnC,qEAAqE;QACrE,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;QAC/B,MAAM,SAAS,GAAiD,EAAE,CAAC;QACnE,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;YACtB,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,MAAM,EAAE;gBACnD,CAAC,EAAE,IAAI,CAAC,CAAC;gBACT,MAAM,EAAE,IAAI,CAAC,MAAM;gBACnB,WAAW,EAAE,IAAI,CAAC,WAAW;aAC9B,CAAC,CAAC;YACH,KAAK,MAAM,EAAE,UAAU,EAAE,CAAC,EAAE,IAAI,IAAI,EAAE,CAAC;gBACrC,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC;oBAAE,SAAS;gBAC7B,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;gBACf,SAAS,CAAC,IAAI,CAAC,EAAE,SAAS,EAAE,CAAC,CAAC,SAAS,EAAE,MAAM,EAAE,CAAC,CAAC,cAAc,EAAE,CAAC,CAAC;gBACrE,IAAI,SAAS,CAAC,MAAM,IAAI,IAAI,CAAC,YAAY;oBAAE,MAAM;YACnD,CAAC;YACD,IAAI,SAAS,CAAC,MAAM,IAAI,IAAI,CAAC,YAAY;gBAAE,MAAM;QACnD,CAAC;QACD,IAAI,SAAS,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QAEtC,IAAI,IAAI,CAAC,IAAI,KAAK,QAAQ,EAAE,CAAC;YAC3B,gEAAgE;YAChE,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;YAC/B,KAAK,MAAM,CAAC,IAAI,SAAS;gBAAE,KAAK,MAAM,CAAC,IAAI,mBAAmB,CAAC,CAAC,CAAC,MAAM,CAAC;oBAAE,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;YACtF,IAAI,IAAI,CAAC,IAAI,GAAG,CAAC,EAAE,CAAC;gBAClB,OAAO;oBACL,6EAA6E;oBAC7E,0DAA0D;oBAC1D,GAAG,CAAC,GAAG,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,EAAE,CAAC;iBAClC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YACf,CAAC;YACD,4EAA4E;QAC9E,CAAC;QAED,yDAAyD;QACzD,OAAO;YACL,6EAA6E;YAC7E,oEAAoE;YACpE,GAAG,SAAS,CAAC,GAAG,CACd,CAAC,CAAC,EAAE,EAAE,CACJ,WAAW,CAAC,CAAC,SAAS,iCAAiC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,YAAY,CAAC,EAAE,CAChG;SACF,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACf,CAAC;IAED,KAAK,CAAC,KAAK;QACT,MAAM,IAAI,CAAC,MAAM,CAAC,KAAK,EAAE,CAAC;IAC5B,CAAC;CACF"}
|
|
@@ -0,0 +1,39 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* SkillOpt v1 rule-judge primitives + task shape.
|
|
3
|
+
*
|
|
4
|
+
* Extracted from skillopt_v1.ts so the memory/causal lever modules can import
|
|
5
|
+
* these without importing the main module (which type-imports the levers back
|
|
6
|
+
* — the old arrangement was a 3-way import cycle). skillopt_v1.ts re-exports
|
|
7
|
+
* everything here, so existing importers are unaffected.
|
|
8
|
+
*/
|
|
9
|
+
export interface RuleCheck {
|
|
10
|
+
op: string;
|
|
11
|
+
arg: unknown;
|
|
12
|
+
}
|
|
13
|
+
export interface RuleJudge {
|
|
14
|
+
kind?: string;
|
|
15
|
+
checks?: RuleCheck[];
|
|
16
|
+
}
|
|
17
|
+
/** Human-readable description of a single check (for causal/reflexion signals). */
|
|
18
|
+
export declare function describeCheck(c: RuleCheck): string;
|
|
19
|
+
export interface CheckResult {
|
|
20
|
+
op: string;
|
|
21
|
+
arg: unknown;
|
|
22
|
+
pass: boolean;
|
|
23
|
+
desc: string;
|
|
24
|
+
}
|
|
25
|
+
/** Per-check pass/fail breakdown for a response (used by the causal lever). */
|
|
26
|
+
export declare function evaluateChecks(judge: RuleJudge, response: string, toolsCalled?: string[]): CheckResult[];
|
|
27
|
+
/** (hard, soft, rationale) for a gbrain-style rule judge. hard=1 iff ALL pass. */
|
|
28
|
+
export declare function scoreRuleJudge(judge: RuleJudge, response: string, toolsCalled?: string[]): {
|
|
29
|
+
hard: number;
|
|
30
|
+
soft: number;
|
|
31
|
+
rationale: string;
|
|
32
|
+
};
|
|
33
|
+
export interface SkillOptV1Task {
|
|
34
|
+
id: string;
|
|
35
|
+
intent: string;
|
|
36
|
+
judge: RuleJudge;
|
|
37
|
+
split: 'train' | 'val' | 'test';
|
|
38
|
+
}
|
|
39
|
+
//# sourceMappingURL=skillopt_v1_rules.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"skillopt_v1_rules.d.ts","sourceRoot":"","sources":["../../../src/eval/live/skillopt_v1_rules.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAIH,MAAM,WAAW,SAAS;IACxB,EAAE,EAAE,MAAM,CAAC;IACX,GAAG,EAAE,OAAO,CAAC;CACd;AACD,MAAM,WAAW,SAAS;IACxB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,SAAS,EAAE,CAAC;CACtB;AA2CD,mFAAmF;AACnF,wBAAgB,aAAa,CAAC,CAAC,EAAE,SAAS,GAAG,MAAM,CAiBlD;AAED,MAAM,WAAW,WAAW;IAC1B,EAAE,EAAE,MAAM,CAAC;IACX,GAAG,EAAE,OAAO,CAAC;IACb,IAAI,EAAE,OAAO,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd;AAED,+EAA+E;AAC/E,wBAAgB,cAAc,CAC5B,KAAK,EAAE,SAAS,EAChB,QAAQ,EAAE,MAAM,EAChB,WAAW,GAAE,MAAM,EAAO,GACzB,WAAW,EAAE,CAOf;AAED,kFAAkF;AAClF,wBAAgB,cAAc,CAC5B,KAAK,EAAE,SAAS,EAChB,QAAQ,EAAE,MAAM,EAChB,WAAW,GAAE,MAAM,EAAO,GACzB;IAAE,IAAI,EAAE,MAAM,CAAC;IAAC,IAAI,EAAE,MAAM,CAAC;IAAC,SAAS,EAAE,MAAM,CAAA;CAAE,CAYnD;AAED,MAAM,WAAW,cAAc;IAC7B,EAAE,EAAE,MAAM,CAAC;IACX,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE,SAAS,CAAC;IACjB,KAAK,EAAE,OAAO,GAAG,KAAK,GAAG,MAAM,CAAC;CACjC"}
|
|
@@ -0,0 +1,91 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* SkillOpt v1 rule-judge primitives + task shape.
|
|
3
|
+
*
|
|
4
|
+
* Extracted from skillopt_v1.ts so the memory/causal lever modules can import
|
|
5
|
+
* these without importing the main module (which type-imports the levers back
|
|
6
|
+
* — the old arrangement was a 3-way import cycle). skillopt_v1.ts re-exports
|
|
7
|
+
* everything here, so existing importers are unaffected.
|
|
8
|
+
*/
|
|
9
|
+
function sectionPresent(response, name) {
|
|
10
|
+
const esc = name.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
|
|
11
|
+
// markdown heading or bold line containing the name (port of judges.py)
|
|
12
|
+
const heading = new RegExp(`^\\s{0,3}(#{1,6}\\s*.*${esc}|\\*\\*.*${esc}.*\\*\\*\\s*:?)\\s*$`, 'im');
|
|
13
|
+
if (heading.test(response || ''))
|
|
14
|
+
return true;
|
|
15
|
+
// "Name:" style label at line start
|
|
16
|
+
const label = new RegExp(`^\\s*${esc}\\s*:`, 'im');
|
|
17
|
+
return label.test(response || '');
|
|
18
|
+
}
|
|
19
|
+
function checkOne(op, arg, response, toolsCalled) {
|
|
20
|
+
const r = response || '';
|
|
21
|
+
switch (op) {
|
|
22
|
+
case 'section_present':
|
|
23
|
+
return sectionPresent(r, String(arg));
|
|
24
|
+
case 'regex':
|
|
25
|
+
try {
|
|
26
|
+
return new RegExp(String(arg)).test(r);
|
|
27
|
+
}
|
|
28
|
+
catch {
|
|
29
|
+
return false;
|
|
30
|
+
}
|
|
31
|
+
case 'max_chars':
|
|
32
|
+
return r.length <= Number(arg);
|
|
33
|
+
case 'min_chars':
|
|
34
|
+
return r.length >= Number(arg);
|
|
35
|
+
case 'contains':
|
|
36
|
+
return r.toLowerCase().includes(String(arg).toLowerCase());
|
|
37
|
+
case 'tool_called': {
|
|
38
|
+
const name = String(arg).toLowerCase();
|
|
39
|
+
if (toolsCalled.some((t) => t.toLowerCase() === name))
|
|
40
|
+
return true;
|
|
41
|
+
// single-shot approximation: an explicit marker in the response
|
|
42
|
+
return new RegExp(`\\btool_call\\s*:\\s*${name}\\b`, 'i').test(r);
|
|
43
|
+
}
|
|
44
|
+
default:
|
|
45
|
+
return true; // unknown op: do not block
|
|
46
|
+
}
|
|
47
|
+
}
|
|
48
|
+
/** Human-readable description of a single check (for causal/reflexion signals). */
|
|
49
|
+
export function describeCheck(c) {
|
|
50
|
+
switch (c.op) {
|
|
51
|
+
case 'section_present':
|
|
52
|
+
return `a "${String(c.arg)}" section`;
|
|
53
|
+
case 'regex':
|
|
54
|
+
return `text matching /${String(c.arg)}/`;
|
|
55
|
+
case 'max_chars':
|
|
56
|
+
return `length <= ${String(c.arg)} chars`;
|
|
57
|
+
case 'min_chars':
|
|
58
|
+
return `length >= ${String(c.arg)} chars`;
|
|
59
|
+
case 'contains':
|
|
60
|
+
return `the text "${String(c.arg)}"`;
|
|
61
|
+
case 'tool_called':
|
|
62
|
+
return `a call to the "${String(c.arg)}" tool`;
|
|
63
|
+
default:
|
|
64
|
+
return `${c.op}=${String(c.arg)}`;
|
|
65
|
+
}
|
|
66
|
+
}
|
|
67
|
+
/** Per-check pass/fail breakdown for a response (used by the causal lever). */
|
|
68
|
+
export function evaluateChecks(judge, response, toolsCalled = []) {
|
|
69
|
+
return (judge?.checks ?? []).map((c) => ({
|
|
70
|
+
op: c.op ?? '',
|
|
71
|
+
arg: c.arg,
|
|
72
|
+
pass: checkOne(c.op ?? '', c.arg, response, toolsCalled),
|
|
73
|
+
desc: describeCheck(c),
|
|
74
|
+
}));
|
|
75
|
+
}
|
|
76
|
+
/** (hard, soft, rationale) for a gbrain-style rule judge. hard=1 iff ALL pass. */
|
|
77
|
+
export function scoreRuleJudge(judge, response, toolsCalled = []) {
|
|
78
|
+
const results = evaluateChecks(judge, response, toolsCalled);
|
|
79
|
+
if (results.length === 0)
|
|
80
|
+
return { hard: 0, soft: 0, rationale: 'no checks' };
|
|
81
|
+
const passed = results.filter((r) => r.pass).length;
|
|
82
|
+
const failed = results.filter((r) => !r.pass).map((r) => `${r.op}=${String(r.arg)}`);
|
|
83
|
+
const soft = passed / results.length;
|
|
84
|
+
const hard = passed === results.length ? 1 : 0;
|
|
85
|
+
return {
|
|
86
|
+
hard,
|
|
87
|
+
soft,
|
|
88
|
+
rationale: hard ? 'all checks passed' : `failed: ${failed.join(', ')}`,
|
|
89
|
+
};
|
|
90
|
+
}
|
|
91
|
+
//# sourceMappingURL=skillopt_v1_rules.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"skillopt_v1_rules.js","sourceRoot":"","sources":["../../../src/eval/live/skillopt_v1_rules.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAaH,SAAS,cAAc,CAAC,QAAgB,EAAE,IAAY;IACpD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,qBAAqB,EAAE,MAAM,CAAC,CAAC;IACxD,wEAAwE;IACxE,MAAM,OAAO,GAAG,IAAI,MAAM,CACxB,yBAAyB,GAAG,YAAY,GAAG,sBAAsB,EACjE,IAAI,CACL,CAAC;IACF,IAAI,OAAO,CAAC,IAAI,CAAC,QAAQ,IAAI,EAAE,CAAC;QAAE,OAAO,IAAI,CAAC;IAC9C,oCAAoC;IACpC,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,QAAQ,GAAG,OAAO,EAAE,IAAI,CAAC,CAAC;IACnD,OAAO,KAAK,CAAC,IAAI,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC;AACpC,CAAC;AAED,SAAS,QAAQ,CAAC,EAAU,EAAE,GAAY,EAAE,QAAgB,EAAE,WAAqB;IACjF,MAAM,CAAC,GAAG,QAAQ,IAAI,EAAE,CAAC;IACzB,QAAQ,EAAE,EAAE,CAAC;QACX,KAAK,iBAAiB;YACpB,OAAO,cAAc,CAAC,CAAC,EAAE,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC;QACxC,KAAK,OAAO;YACV,IAAI,CAAC;gBACH,OAAO,IAAI,MAAM,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACzC,CAAC;YAAC,MAAM,CAAC;gBACP,OAAO,KAAK,CAAC;YACf,CAAC;QACH,KAAK,WAAW;YACd,OAAO,CAAC,CAAC,MAAM,IAAI,MAAM,CAAC,GAAG,CAAC,CAAC;QACjC,KAAK,WAAW;YACd,OAAO,CAAC,CAAC,MAAM,IAAI,MAAM,CAAC,GAAG,CAAC,CAAC;QACjC,KAAK,UAAU;YACb,OAAO,CAAC,CAAC,WAAW,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC;QAC7D,KAAK,aAAa,CAAC,CAAC,CAAC;YACnB,MAAM,IAAI,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,WAAW,EAAE,CAAC;YACvC,IAAI,WAAW,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,KAAK,IAAI,CAAC;gBAAE,OAAO,IAAI,CAAC;YACnE,gEAAgE;YAChE,OAAO,IAAI,MAAM,CAAC,wBAAwB,IAAI,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACpE,CAAC;QACD;YACE,OAAO,IAAI,CAAC,CAAC,2BAA2B;IAC5C,CAAC;AACH,CAAC;AAED,mFAAmF;AACnF,MAAM,UAAU,aAAa,CAAC,CAAY;IACxC,QAAQ,CAAC,CAAC,EAAE,EAAE,CAAC;QACb,KAAK,iBAAiB;YACpB,OAAO,MAAM,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,WAAW,CAAC;QACxC,KAAK,OAAO;YACV,OAAO,kBAAkB,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC;QAC5C,KAAK,WAAW;YACd,OAAO,aAAa,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,QAAQ,CAAC;QAC5C,KAAK,WAAW;YACd,OAAO,aAAa,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,QAAQ,CAAC;QAC5C,KAAK,UAAU;YACb,OAAO,aAAa,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC;QACvC,KAAK,aAAa;YAChB,OAAO,kBAAkB,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,QAAQ,CAAC;QACjD;YACE,OAAO,GAAG,CAAC,CAAC,EAAE,IAAI,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;IACtC,CAAC;AACH,CAAC;AASD,+EAA+E;AAC/E,MAAM,UAAU,cAAc,CAC5B,KAAgB,EAChB,QAAgB,EAChB,cAAwB,EAAE;IAE1B,OAAO,CAAC,KAAK,EAAE,MAAM,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACvC,EAAE,EAAE,CAAC,CAAC,EAAE,IAAI,EAAE;QACd,GAAG,EAAE,CAAC,CAAC,GAAG;QACV,IAAI,EAAE,QAAQ,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,QAAQ,EAAE,WAAW,CAAC;QACxD,IAAI,EAAE,aAAa,CAAC,CAAC,CAAC;KACvB,CAAC,CAAC,CAAC;AACN,CAAC;AAED,kFAAkF;AAClF,MAAM,UAAU,cAAc,CAC5B,KAAgB,EAChB,QAAgB,EAChB,cAAwB,EAAE;IAE1B,MAAM,OAAO,GAAG,cAAc,CAAC,KAAK,EAAE,QAAQ,EAAE,WAAW,CAAC,CAAC;IAC7D,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,SAAS,EAAE,WAAW,EAAE,CAAC;IAC9E,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC;IACpD,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IACrF,MAAM,IAAI,GAAG,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC;IACrC,MAAM,IAAI,GAAG,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC/C,OAAO;QACL,IAAI;QACJ,IAAI;QACJ,SAAS,EAAE,IAAI,CAAC,CAAC,CAAC,mBAAmB,CAAC,CAAC,CAAC,WAAW,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;KACvE,CAAC;AACJ,CAAC"}
|
|
@@ -0,0 +1,138 @@
|
|
|
1
|
+
import { createStep, type Task } from '../../types/index.js';
|
|
2
|
+
import type { EvalTask } from '../types.js';
|
|
3
|
+
import type { ExtendedLiveBenchmark, BenchmarkVerdict } from './live-benchmark.js';
|
|
4
|
+
import type { BenchmarkDelegate } from './live-runner.js';
|
|
5
|
+
import { DockerOrchestrator } from './docker.js';
|
|
6
|
+
/**
|
|
7
|
+
* SkillsBench — one-shot agentic tasks across multiple domains.
|
|
8
|
+
*
|
|
9
|
+
* The community standard for measuring skill quality: each task has a Docker
|
|
10
|
+
* container, a task instruction, and a ground-truth test suite. The agent
|
|
11
|
+
* gets one shot (with optional skill injection), and the test suite determines
|
|
12
|
+
* pass/fail. No LLM judge.
|
|
13
|
+
*
|
|
14
|
+
* This adapter loads tasks from a local SkillsBench repo (cloned by the user),
|
|
15
|
+
* builds Docker images per-task on setup, and runs agent + tests per task.
|
|
16
|
+
*
|
|
17
|
+
* Repo structure (Harbor format):
|
|
18
|
+
* tasks/<task-id>/
|
|
19
|
+
* task.toml — metadata (category, difficulty, timeouts)
|
|
20
|
+
* instruction.md — task prompt
|
|
21
|
+
* environment/
|
|
22
|
+
* Dockerfile — container definition
|
|
23
|
+
* tests/ — test suite (run inside container)
|
|
24
|
+
*
|
|
25
|
+
* Environment variables:
|
|
26
|
+
* SKILLSBENCH_REPO — path to the cloned SkillsBench repository
|
|
27
|
+
* SKILLSBENCH_TIMEOUT — per-task timeout in ms (default 300_000)
|
|
28
|
+
*/
|
|
29
|
+
interface SkillsBenchTaskDef {
|
|
30
|
+
id: string;
|
|
31
|
+
/** Category from task.toml metadata (e.g. "industrial-physical-systems"). */
|
|
32
|
+
domain: string;
|
|
33
|
+
instruction: string;
|
|
34
|
+
dockerfilePath: string;
|
|
35
|
+
testCommand: string[];
|
|
36
|
+
workdir: string;
|
|
37
|
+
skillDir: string;
|
|
38
|
+
/** Difficulty from task.toml (e.g. "hard", "medium"). */
|
|
39
|
+
difficulty?: string;
|
|
40
|
+
/** Tags from task.toml for searchability. */
|
|
41
|
+
tags?: string[];
|
|
42
|
+
/** Absolute path to the task directory on host. */
|
|
43
|
+
taskDir: string;
|
|
44
|
+
}
|
|
45
|
+
export declare class SkillsBenchBenchmark implements ExtendedLiveBenchmark {
|
|
46
|
+
private readonly opts;
|
|
47
|
+
readonly name = "skillsbench";
|
|
48
|
+
readonly interactive = false;
|
|
49
|
+
readonly preserveOrder = false;
|
|
50
|
+
readonly playbookMinTrajectories = 3;
|
|
51
|
+
private docker;
|
|
52
|
+
private repoPath;
|
|
53
|
+
private taskDefs;
|
|
54
|
+
private builtDomainImages;
|
|
55
|
+
private traceDir;
|
|
56
|
+
constructor(opts?: {
|
|
57
|
+
repoPath?: string;
|
|
58
|
+
timeoutMs?: number;
|
|
59
|
+
});
|
|
60
|
+
setup(): Promise<void>;
|
|
61
|
+
teardown(): Promise<void>;
|
|
62
|
+
loadTasks(opts: {
|
|
63
|
+
count: number;
|
|
64
|
+
seed: number;
|
|
65
|
+
}): Promise<EvalTask[]>;
|
|
66
|
+
verify(_task: Task, _output: string): BenchmarkVerdict;
|
|
67
|
+
getDocker(): DockerOrchestrator;
|
|
68
|
+
getTaskDef(taskDefId: string): SkillsBenchTaskDef | undefined;
|
|
69
|
+
/**
|
|
70
|
+
* Get the Docker image for a task. Falls back to domain-level lookup
|
|
71
|
+
* for backwards compatibility with domain-nested layouts.
|
|
72
|
+
*/
|
|
73
|
+
getTaskImage(taskDefId: string): string | undefined;
|
|
74
|
+
getDomainImage(domain: string): string | undefined;
|
|
75
|
+
/**
|
|
76
|
+
* Build the Docker image for a single task on demand.
|
|
77
|
+
* Returns the image tag, or undefined if no Dockerfile exists.
|
|
78
|
+
*/
|
|
79
|
+
ensureImage(taskDefId: string): Promise<string | undefined>;
|
|
80
|
+
getTimeoutMs(): number;
|
|
81
|
+
getTraceDir(): string | undefined;
|
|
82
|
+
/**
|
|
83
|
+
* Load task definitions from the SkillsBench repo.
|
|
84
|
+
*
|
|
85
|
+
* Supports the actual Harbor format:
|
|
86
|
+
* tasks/<task-id>/
|
|
87
|
+
* task.toml — metadata
|
|
88
|
+
* instruction.md — task prompt
|
|
89
|
+
* environment/Dockerfile — container definition
|
|
90
|
+
* tests/ — test suite
|
|
91
|
+
*
|
|
92
|
+
* Also supports legacy nested structure for backwards compatibility:
|
|
93
|
+
* tasks/<domain>/<task-id>/instruction.md
|
|
94
|
+
*/
|
|
95
|
+
private loadManifest;
|
|
96
|
+
/**
|
|
97
|
+
* Recursively scan for task directories. A directory is a task if it contains
|
|
98
|
+
* instruction.md (and optionally task.toml for Harbor metadata).
|
|
99
|
+
*/
|
|
100
|
+
private scanTaskDir;
|
|
101
|
+
private resolveDockerfile;
|
|
102
|
+
private resolveTestCommand;
|
|
103
|
+
/**
|
|
104
|
+
* Infer domain from task ID path (e.g. "web-scraping/task-1" → "web-scraping").
|
|
105
|
+
* For flat IDs, uses the task ID itself.
|
|
106
|
+
*/
|
|
107
|
+
private inferDomain;
|
|
108
|
+
/**
|
|
109
|
+
* Minimal task.toml parser for the subset used by Harbor benchmarks.
|
|
110
|
+
* Handles strings, numbers, booleans, arrays, and [section] headers.
|
|
111
|
+
*/
|
|
112
|
+
private parseTaskToml;
|
|
113
|
+
private parseTomlValue;
|
|
114
|
+
private shuffle;
|
|
115
|
+
delegateFactory: (_benchmark: ExtendedLiveBenchmark, opts?: {
|
|
116
|
+
model?: string;
|
|
117
|
+
timeoutMs?: number;
|
|
118
|
+
provider?: string;
|
|
119
|
+
awsRegion?: string;
|
|
120
|
+
}) => BenchmarkDelegate;
|
|
121
|
+
}
|
|
122
|
+
export declare function createSkillsBenchBenchmark(opts?: {
|
|
123
|
+
repoPath?: string;
|
|
124
|
+
timeoutMs?: number;
|
|
125
|
+
}): SkillsBenchBenchmark;
|
|
126
|
+
/**
|
|
127
|
+
* Map the per-step command/output records collected during the multi-turn loop
|
|
128
|
+
* into real Step objects so the delegate result carries a full trajectory.
|
|
129
|
+
*
|
|
130
|
+
* Factored as a pure function for unit-testing without Docker.
|
|
131
|
+
*/
|
|
132
|
+
export declare function mapSkillsBenchSteps(steps: Array<{
|
|
133
|
+
command: string;
|
|
134
|
+
exitCode: number;
|
|
135
|
+
output: string;
|
|
136
|
+
}>, initialObs: string): ReturnType<typeof createStep>[];
|
|
137
|
+
export {};
|
|
138
|
+
//# sourceMappingURL=skillsbench.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"skillsbench.d.ts","sourceRoot":"","sources":["../../../src/eval/live/skillsbench.ts"],"names":[],"mappings":"AAGA,OAAO,EAGL,UAAU,EAGV,KAAK,IAAI,EACV,MAAM,sBAAsB,CAAC;AAE9B,OAAO,KAAK,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAC;AAC5C,OAAO,KAAK,EAAE,qBAAqB,EAAE,gBAAgB,EAAE,MAAM,qBAAqB,CAAC;AACnF,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,kBAAkB,CAAC;AAC1D,OAAO,EAAwB,kBAAkB,EAA4B,MAAM,aAAa,CAAC;AAIjG;;;;;;;;;;;;;;;;;;;;;;GAsBG;AAEH,UAAU,kBAAkB;IAC1B,EAAE,EAAE,MAAM,CAAC;IACX,6EAA6E;IAC7E,MAAM,EAAE,MAAM,CAAC;IACf,WAAW,EAAE,MAAM,CAAC;IACpB,cAAc,EAAE,MAAM,CAAC;IACvB,WAAW,EAAE,MAAM,EAAE,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,MAAM,CAAC;IACjB,yDAAyD;IACzD,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,6CAA6C;IAC7C,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC;IAChB,mDAAmD;IACnD,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,qBAAa,oBAAqB,YAAW,qBAAqB;IAYpD,OAAO,CAAC,QAAQ,CAAC,IAAI;IAXjC,QAAQ,CAAC,IAAI,iBAAiB;IAC9B,QAAQ,CAAC,WAAW,SAAS;IAC7B,QAAQ,CAAC,aAAa,SAAS;IAC/B,QAAQ,CAAC,uBAAuB,KAAK;IAErC,OAAO,CAAC,MAAM,CAAqB;IACnC,OAAO,CAAC,QAAQ,CAAS;IACzB,OAAO,CAAC,QAAQ,CAA4B;IAC5C,OAAO,CAAC,iBAAiB,CAA6B;IACtD,OAAO,CAAC,QAAQ,CAAqB;gBAER,IAAI,GAAE;QAAE,QAAQ,CAAC,EAAE,MAAM,CAAC;QAAC,SAAS,CAAC,EAAE,MAAM,CAAA;KAAO;IAK3E,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;IAatB,QAAQ,IAAI,OAAO,CAAC,IAAI,CAAC;IAOzB,SAAS,CAAC,IAAI,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,GAAG,OAAO,CAAC,QAAQ,EAAE,CAAC;IAyB3E,MAAM,CAAC,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,GAAG,gBAAgB;IAItD,SAAS,IAAI,kBAAkB;IAI/B,UAAU,CAAC,SAAS,EAAE,MAAM,GAAG,kBAAkB,GAAG,SAAS;IAI7D;;;OAGG;IACH,YAAY,CAAC,SAAS,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAInD,cAAc,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAIlD;;;OAGG;IACG,WAAW,CAAC,SAAS,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,GAAG,SAAS,CAAC;IAejE,YAAY,IAAI,MAAM;IAItB,WAAW,IAAI,MAAM,GAAG,SAAS;IAIjC;;;;;;;;;;;;OAYG;IACH,OAAO,CAAC,YAAY;IAWpB;;;OAGG;IACH,OAAO,CAAC,WAAW;IAoDnB,OAAO,CAAC,iBAAiB;IAYzB,OAAO,CAAC,kBAAkB;IAyB1B;;;OAGG;IACH,OAAO,CAAC,WAAW;IAKnB;;;OAGG;IACH,OAAO,CAAC,aAAa;IAiCrB,OAAO,CAAC,cAAc;IAiBtB,OAAO,CAAC,OAAO;IAUf,eAAe,GACb,YAAY,qBAAqB,EACjC,OAAO;QAAE,KAAK,CAAC,EAAE,MAAM,CAAC;QAAC,SAAS,CAAC,EAAE,MAAM,CAAC;QAAC,QAAQ,CAAC,EAAE,MAAM,CAAC;QAAC,SAAS,CAAC,EAAE,MAAM,CAAA;KAAE,KACnF,iBAAiB,CAOlB;CACH;AAED,wBAAgB,0BAA0B,CAAC,IAAI,CAAC,EAAE;IAChD,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB,GAAG,oBAAoB,CAEvB;AAyYD;;;;;GAKG;AACH,wBAAgB,mBAAmB,CACjC,KAAK,EAAE,KAAK,CAAC;IAAE,OAAO,EAAE,MAAM,CAAC;IAAC,QAAQ,EAAE,MAAM,CAAC;IAAC,MAAM,EAAE,MAAM,CAAA;CAAE,CAAC,EACnE,UAAU,EAAE,MAAM,GACjB,UAAU,CAAC,OAAO,UAAU,CAAC,EAAE,CAqBjC"}
|