Change detail
1 change across 1 operation
- POST/chat-sessions/messages
- info
added the optional property
to the response with the status
- info
- Let goals choose their judge model over the API and at swarm creation (#6524)6Docs4a19ae72026-10-19 12:33
Change detail
6 changes across 4 operations
- PATCH/projects/{projectId}/goals/{goalId}
- info
added the new optional request property
- info
added the optional property
to the response with the status
- info
- POST/projects/{projectId}/goals
- info
added the new optional request property
- info
added the optional property
to the response with the status
- info
- GET/projects/{projectId}/goals/{goalId}
- info
added the optional property
to the response with the status
- info
- GET/projects/{projectId}/goals
- info
added the optional property
//to the response with the status
- info
Change detail
1 change across 1 operation
- POST/projects/{projectId}/eval-ingest/pay-iteration
- info
endpoint added
- info
Change detail
1 change across 1 operation
- POST/projects/{projectId}/eval-ingest/start-run
- info
endpoint added
- info
Change detail
2 changes across 1 operation
- POST/projects/{projectId}/model-leases
- info
added the new optional request property
- info
added the new optional request property
- info
- Muse directory readiness on the Conformance page, API, CLI and agents (#6056)82Docs76a9d012026-10-07 22:31
- Evals: rerun a run's failed cases over the API and CLI; keep subset reruns out of client aggregates (#5890)8Docsd27c3aa2026-10-05 20:05
- Record eval trials our infrastructure failed instead of scoring them (#5893)1Docs49b0f942026-10-05 09:00
- Model selection: effort UI, compare by effort, routing by saved selection, usage by effort (#5809)84114Docs1b5edf12026-10-04 20:18
- Reasoning effort: adapters, chat, UI, API, CLI and SDK (effort plan, PRs 7–14) (#5760)34Docsf9dfb602026-10-04 19:39
- Add send_feedback: POST /v1/feedback, SDK operation and MCP tool (#5707)1Docs9094f762026-09-29 03:38
- Evals: toolInputMatches + toolResultMatches (rename before release, path as JSON Pointer) (#5680)16418Docs4c077362026-09-28 09:00
- Evals: toolArgumentsMatch — same-call regex check over tool arguments (#5674)1125Docsd9d4fbd2026-09-28 07:16
- Public API vocabulary consolidation: scenario → study, journey → goal, wave → swarm run (#5429)413114Docsafd161a2026-09-22 23:36
- Swarm findings: say the same thing whichever order the rows arrive in (#5359)12Docs2d59e6f2026-09-21 02:10
- Swarm findings contract: recorded signals, persona account, verification counts (#5340)1218Docscc8f06a2026-09-19 19:36
- Expose typed swarm findings across SDK, API, MCP, and Inspector (#5281)12Docs4311e1f2026-09-18 04:03
- Report swarm execution, goal grading, and evidence independently (#5248)7Docsb99eed62026-09-17 06:26
- feat: run SDK tests with latest saved clients and show version tooltips (#5175)10Docsf1985632026-09-16 18:24
- Evaluate: findings panel and the empty-step engine fix (lean rework of #5066) (#5112)24Docs4e68dc22026-09-15 20:28
- Evals: one grading policy, one vocabulary (G0a, G0c, G1, G2, G3) (#5070)25Docsbbfb7d22026-09-15 11:30
- v1: read x-mcpjam-eval-vocabulary, and advertise the vocabulary block (A0) (#5050)1Docs4b839442026-09-14 21:38
- Unpublish: document ?scenarioId=, and stop a taken fallback name dead-ending (#5025)1Docs6c09a662026-09-14 04:37
- Ship enterprise eval SDK execution, reporting, and evaluator tooling (#5022)18Docs20bf8ac2026-09-14 01:42
- feat(evals): persist standard checks and report discovery assertions through UVC (#5021)1515Docsa108aa42026-09-13 23:36
- Reject a malformed eval id at the v1 boundary instead of paging on it (#4358)15Docs660da592026-09-12 09:17
- Evaluate: case workspace, Markdown case import, and Describe-only agent (#4867)7Docs0d016d82026-09-10 07:00
- Friction signals per trial, and the suspected condition behind them (#4796)4Docs9272dd92026-09-08 19:53