@tangle-network/agent-eval 0.117.1 → 0.118.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +23 -0
- package/dist/analyst/index.d.ts +2772 -21
- package/dist/analyst/index.js +7 -6
- package/dist/analyst/index.js.map +1 -1
- package/dist/belief-state/index.d.ts +706 -10
- package/dist/belief-state/index.js +2 -1
- package/dist/belief-state/index.js.map +1 -1
- package/dist/benchmarks/index.d.ts +958 -14
- package/dist/benchmarks/index.js +12 -10
- package/dist/builder-eval/index.d.ts +449 -4
- package/dist/builder-eval/index.js +4 -3
- package/dist/builder-eval/index.js.map +1 -1
- package/dist/campaign/index.d.ts +4275 -73
- package/dist/campaign/index.js +12 -10
- package/dist/{chunk-VF3XSYTI.js → chunk-33JA4TFA.js} +6 -6
- package/dist/{chunk-4JLWXDYA.js → chunk-3EHHMC6E.js} +2 -2
- package/dist/{chunk-CCZIVI3F.js → chunk-BFW56GTT.js} +2 -2
- package/dist/{chunk-YZPO4UHR.js → chunk-DP3BBMYJ.js} +100 -143
- package/dist/chunk-DP3BBMYJ.js.map +1 -0
- package/dist/{chunk-E4BUPP7Z.js → chunk-HKUCJ437.js} +38 -66
- package/dist/chunk-HKUCJ437.js.map +1 -0
- package/dist/{chunk-HZHNRYHK.js → chunk-K6N6XJJX.js} +2 -2
- package/dist/chunk-KSDQVPLR.js +286 -0
- package/dist/chunk-KSDQVPLR.js.map +1 -0
- package/dist/chunk-MA6HLL3S.js +65 -0
- package/dist/chunk-MA6HLL3S.js.map +1 -0
- package/dist/{chunk-MGEHEHSN.js → chunk-OIIMMLRB.js} +11 -11
- package/dist/{chunk-DXZRATT5.js → chunk-OYZAPX5G.js} +3 -3
- package/dist/chunk-PXE2VKMX.js +140 -0
- package/dist/chunk-PXE2VKMX.js.map +1 -0
- package/dist/{chunk-JSJZ4PJ6.js → chunk-Q442S5AS.js} +17 -17
- package/dist/{chunk-ODVOOEWQ.js → chunk-QBRSJK47.js} +2 -2
- package/dist/{chunk-S2F4J57L.js → chunk-QKEGNI5B.js} +77 -32
- package/dist/chunk-QKEGNI5B.js.map +1 -0
- package/dist/{chunk-5UF54T55.js → chunk-S3UZOQ5Y.js} +34 -7
- package/dist/chunk-S3UZOQ5Y.js.map +1 -0
- package/dist/{chunk-FQNLDL4D.js → chunk-SVH2ANFD.js} +136 -4
- package/dist/chunk-SVH2ANFD.js.map +1 -0
- package/dist/{chunk-GQCZRZ7L.js → chunk-U5CHZ5M3.js} +9 -9
- package/dist/{chunk-TVVP3ZZQ.js → chunk-VQMK5FMP.js} +2 -1
- package/dist/{chunk-TVVP3ZZQ.js.map → chunk-VQMK5FMP.js.map} +1 -1
- package/dist/chunk-WDHBCA3M.js +31 -0
- package/dist/chunk-WDHBCA3M.js.map +1 -0
- package/dist/{chunk-HQPHZGL6.js → chunk-YLMUS4MM.js} +9 -9
- package/dist/{chunk-LQUTGLOZ.js → chunk-ZET2UAYW.js} +15 -65
- package/dist/chunk-ZET2UAYW.js.map +1 -0
- package/dist/contract/index.d.ts +4012 -38
- package/dist/contract/index.js +56 -29
- package/dist/contract/index.js.map +1 -1
- package/dist/control.d.ts +1013 -9
- package/dist/control.js +4 -3
- package/dist/fuzz.d.ts +194 -4
- package/dist/fuzz.js +3 -3
- package/dist/hosted/index.d.ts +498 -17
- package/dist/index.d.ts +10982 -1286
- package/dist/index.js +97 -80
- package/dist/index.js.map +1 -1
- package/dist/matrix/index.d.ts +139 -4
- package/dist/meta-eval/index.d.ts +862 -15
- package/dist/meta-eval/index.js +2 -1
- package/dist/meta-eval/index.js.map +1 -1
- package/dist/multishot/index.d.ts +214 -14
- package/dist/openapi.json +1 -1
- package/dist/pipelines/index.d.ts +392 -7
- package/dist/pipelines/index.js +5 -3
- package/dist/pipelines/index.js.map +1 -1
- package/dist/reporting.d.ts +1277 -17
- package/dist/rl.d.ts +2359 -28
- package/dist/rl.js +6 -5
- package/dist/rl.js.map +1 -1
- package/dist/storyboard/index.d.ts +86 -1
- package/dist/trace-attributes.d.ts +16 -0
- package/dist/trace-attributes.js +32 -0
- package/dist/trace-attributes.js.map +1 -0
- package/dist/traces.d.ts +1970 -697
- package/dist/traces.js +49 -32
- package/dist/wire/index.d.ts +655 -9
- package/docs/insight-report.md +44 -0
- package/package.json +9 -3
- package/dist/adversarial-B7loGVVX.d.ts +0 -19
- package/dist/analyst-C8HHvfJp.d.ts +0 -88
- package/dist/analyze-runs--2x39HZ7.d.ts +0 -81
- package/dist/baseline-DKq3gJpP.d.ts +0 -141
- package/dist/calibration-C8MTS7cw.d.ts +0 -101
- package/dist/chunk-5UF54T55.js.map +0 -1
- package/dist/chunk-E4BUPP7Z.js.map +0 -1
- package/dist/chunk-FQNLDL4D.js.map +0 -1
- package/dist/chunk-LQUTGLOZ.js.map +0 -1
- package/dist/chunk-S2F4J57L.js.map +0 -1
- package/dist/chunk-YZPO4UHR.js.map +0 -1
- package/dist/code-agent-session-CjZsVd19.d.ts +0 -87
- package/dist/control-6vuGfmDH.d.ts +0 -258
- package/dist/cost-ledger-DWy3XdJc.d.ts +0 -183
- package/dist/dataset-NENEzRgk.d.ts +0 -115
- package/dist/default-registry-DaK8b3fv.d.ts +0 -155
- package/dist/emitter-CjD7vUwv.d.ts +0 -122
- package/dist/errors-oeQrLqXC.d.ts +0 -74
- package/dist/failure-cluster-DOAcSJ87.d.ts +0 -76
- package/dist/feedback-trajectory-BUnM58xL.d.ts +0 -348
- package/dist/gepa-eESocoDi.d.ts +0 -642
- package/dist/index-PdX4VnPA.d.ts +0 -423
- package/dist/insight-report-DY4nDW9Q.d.ts +0 -310
- package/dist/integrity-DqlBiLyK.d.ts +0 -81
- package/dist/judge-calibration-7C-IDmKr.d.ts +0 -145
- package/dist/kind-factory-ClZmO25A.d.ts +0 -171
- package/dist/llm-client-qoDd18Qz.d.ts +0 -289
- package/dist/multi-layer-verifier-BsqKuLyN.d.ts +0 -150
- package/dist/off-policy-DiwuKKg7.d.ts +0 -132
- package/dist/outcome-store-rnXLEqSn.d.ts +0 -63
- package/dist/policy-edit-wG9uFEFm.d.ts +0 -455
- package/dist/pre-registration-BWQhJ3vz.d.ts +0 -761
- package/dist/provenance-DpjwyseI.d.ts +0 -541
- package/dist/query-CF7PG61p.d.ts +0 -35
- package/dist/raw-provider-sink-C46HDghv.d.ts +0 -132
- package/dist/release-report-C8G2i5Xi.d.ts +0 -236
- package/dist/researcher-C8XyxQsu.d.ts +0 -387
- package/dist/rubric-predictive-validity-p49lLVrE.d.ts +0 -105
- package/dist/run-record-BDH49H2E.d.ts +0 -360
- package/dist/runtime-trajectory-DGBIUt4B.d.ts +0 -49
- package/dist/schema-B3Q3l9Z_.d.ts +0 -201
- package/dist/semantic-concept-judge-CXnPEJbf.d.ts +0 -723
- package/dist/sequential-5iSVfzl2.d.ts +0 -139
- package/dist/series-convergence-D5OWMBg6.d.ts +0 -33
- package/dist/statistics-KUnG73jH.d.ts +0 -494
- package/dist/storage-DrX3v_5B.d.ts +0 -50
- package/dist/store-C1YxJDEK.d.ts +0 -248
- package/dist/store-DGqD0Pyo.d.ts +0 -116
- package/dist/summary-report-C5bKFfm-.d.ts +0 -445
- package/dist/test-graded-scenario-B0ybnPY7.d.ts +0 -166
- package/dist/types-BSw1rOUB.d.ts +0 -634
- package/dist/types-BUxNaJ8c.d.ts +0 -108
- package/dist/types-BkfcQnxV.d.ts +0 -313
- package/dist/verdict-C9MlYujm.d.ts +0 -35
- /package/dist/{chunk-VF3XSYTI.js.map → chunk-33JA4TFA.js.map} +0 -0
- /package/dist/{chunk-4JLWXDYA.js.map → chunk-3EHHMC6E.js.map} +0 -0
- /package/dist/{chunk-CCZIVI3F.js.map → chunk-BFW56GTT.js.map} +0 -0
- /package/dist/{chunk-HZHNRYHK.js.map → chunk-K6N6XJJX.js.map} +0 -0
- /package/dist/{chunk-MGEHEHSN.js.map → chunk-OIIMMLRB.js.map} +0 -0
- /package/dist/{chunk-DXZRATT5.js.map → chunk-OYZAPX5G.js.map} +0 -0
- /package/dist/{chunk-JSJZ4PJ6.js.map → chunk-Q442S5AS.js.map} +0 -0
- /package/dist/{chunk-ODVOOEWQ.js.map → chunk-QBRSJK47.js.map} +0 -0
- /package/dist/{chunk-GQCZRZ7L.js.map → chunk-U5CHZ5M3.js.map} +0 -0
- /package/dist/{chunk-HQPHZGL6.js.map → chunk-YLMUS4MM.js.map} +0 -0
package/CHANGELOG.md
CHANGED
|
@@ -4,6 +4,29 @@ All notable changes to `@tangle-network/agent-eval` and its sibling `agent-eval-
|
|
|
4
4
|
|
|
5
5
|
---
|
|
6
6
|
|
|
7
|
+
## [0.118.0] — 2026-07-13 — complete execution accounting
|
|
8
|
+
|
|
9
|
+
### Added
|
|
10
|
+
|
|
11
|
+
- `InsightReport.execution` reports duration, optional queue time, direct input, output, reasoning, cache-read, and cache-write tokens, model cohorts, model-call coverage, explicit failure counts, and separately labeled orchestration aggregates from the same `RunRecord[]` passed to `analyzeRuns()`.
|
|
12
|
+
- `summarizeExecution()` returns those execution facts and cost provenance without interpreting task quality.
|
|
13
|
+
- `RunTokenUsage.reasoning` preserves the reasoning subset of normalized output, and `RunTokenUsage.cacheWrite` preserves provider cache creation separately from cache reads.
|
|
14
|
+
|
|
15
|
+
### Changed
|
|
16
|
+
|
|
17
|
+
- Trace capture, every OTLP exporter, OTLP intake, and code-agent session intake preserve reasoning, cache reads, and cache writes separately.
|
|
18
|
+
- Both OTLP intake paths use one field-by-field reconciliation rule for nested model-call wrappers, preserving complementary parent data without double-counting complete child data.
|
|
19
|
+
- Code-agent intake uses the shared provider-usage parser, including OpenCode's nested `cache.read` and `cache.write` fields and OpenAI-compatible token-detail objects.
|
|
20
|
+
- OTLP-derived run records explicitly label complete USD as observed, model-priced USD as estimated, and missing or partial USD as uncaptured instead of relying on a zero-value inference.
|
|
21
|
+
- Usage parsing reuses `@tangle-network/agent-core` token vocabulary and SSE framing, preserves agent-eval-specific reasoning and cache-write details, reconciles cumulative streams by default, and accepts explicit delta mode through `captureFetchToRawSink({ sseUsageMode })`.
|
|
22
|
+
- Run-record validation rejects negative execution measurements and unknown failure classes, and OTLP intake rejects duplicate span identities instead of corrupting totals.
|
|
23
|
+
- Declaration bundles build sequentially and package verification compiles a strict Node consumer, preserving the public subpath types without concurrent declaration workers exhausting memory.
|
|
24
|
+
- `@tangle-network/agent-interface` is updated to `0.26.x`.
|
|
25
|
+
|
|
26
|
+
### Breaking
|
|
27
|
+
|
|
28
|
+
- `InsightReport.execution`, `CodeAgentSessionMetrics.reasoningTokens`, and `CodeAgentSessionMetrics.cacheWriteTokens` are required fields on newly constructed objects.
|
|
29
|
+
|
|
7
30
|
## [0.117.1] — 2026-07-13 — retry-safe code-candidate cleanup
|
|
8
31
|
|
|
9
32
|
### Fixed
|