diff --git a/docs/whats-new.md b/docs/whats-new.md index 0710aa1..ff7ceb6 100644 --- a/docs/whats-new.md +++ b/docs/whats-new.md @@ -1,13 +1,17 @@ # What's New > *Auto-generated digest of recent changes across the OpenAdapt ecosystem.* -> *Last updated: 2026-09-02 16:27 UTC* +> *Last updated: 2026-09-03 01:45 UTC* ## OpenAdapt +- [copy: two authoring modes, person-click and watch-and-admit](https://github.com/OpenAdaptAI/OpenAdapt/pull/1141) (#1141) — merged + +- [docs: name seven until-revoked target admissions in the README](https://github.com/OpenAdaptAI/OpenAdapt/pull/1140) (#1140) — merged + - [docs: lead with flow tutorial as the VERIFIED first command](https://github.com/OpenAdaptAI/OpenAdapt/pull/1139) (#1139) — merged - [docs: front the agent-callable path](https://github.com/OpenAdaptAI/OpenAdapt/pull/1137) (#1137) — merged @@ -36,23 +40,15 @@ - [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/OpenAdapt/pull/1124) (#1124) — merged -- [chore(release): reconcile platform manifest substrate projection](https://github.com/OpenAdaptAI/OpenAdapt/pull/1121) (#1121) — merged - - [ci: gate launcher release publication](https://github.com/OpenAdaptAI/OpenAdapt/pull/1120) (#1120) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/OpenAdapt/pull/1119) (#1119) — merged - -- [ci: bind launcher releases to the release app](https://github.com/OpenAdaptAI/OpenAdapt/pull/1118) (#1118) — merged - -- [Replace the legacy GitBook link](https://github.com/OpenAdaptAI/OpenAdapt/pull/1117) (#1117) — merged - -- [Derive launcher lifecycle from signed admissions](https://github.com/OpenAdaptAI/OpenAdapt/pull/1116) (#1116) — merged - ## openadapt-flow +- [fix(reward): refuse a bundle whose required effects ignore an identity key](https://github.com/OpenAdaptAI/openadapt-flow/pull/456) (#456) — merged + - [chore: release 1.35.0](https://github.com/OpenAdaptAI/openadapt-flow/pull/454) (#454) — merged - [fix(reward): trainer adapters live in openadapt_evals.reward; flow keeps the worker and the client](https://github.com/OpenAdaptAI/openadapt-flow/pull/453) (#453) — merged @@ -91,8 +87,6 @@ - [feat(visualize): render composed parents without merging ProgramGraphs](https://github.com/OpenAdaptAI/openadapt-flow/pull/433) (#433) — merged -- [feat(compiler): admit REST bindings only after held-out Effect confirmation](https://github.com/OpenAdaptAI/openadapt-flow/pull/432) (#432) — merged - ## openadapt-desktop @@ -136,13 +130,15 @@ - [Guide the first real workflow run](https://github.com/OpenAdaptAI/openadapt-desktop/pull/135) (#135) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/openadapt-desktop/pull/134) (#134) — merged +- [feat: route hosted runs through Flow's governed adapter](https://github.com/OpenAdaptAI/openadapt-desktop/pull/130) (#130) — merged ## openadapt-agent +- [copy: two authoring modes, person-click and watch-and-admit](https://github.com/OpenAdaptAI/openadapt-agent/pull/48) (#48) — merged + - [docs: lead pip install with `openadapt flow tutorial`](https://github.com/OpenAdaptAI/openadapt-agent/pull/47) (#47) — merged - [docs: rewrite README as a calling-agent landing page](https://github.com/OpenAdaptAI/openadapt-agent/pull/46) (#46) — merged @@ -163,10 +159,6 @@ - [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/openadapt-agent/pull/36) (#36) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/openadapt-agent/pull/35) (#35) — merged - -- [Derive Agent state from release admission](https://github.com/OpenAdaptAI/openadapt-agent/pull/33) (#33) — merged - - [chore(deps): Bump astral-sh/setup-uv from 9.0.0 to 10.0.1](https://github.com/OpenAdaptAI/openadapt-agent/pull/32) (#32) — merged - [chore(deps): Bump the codeql-action group with 2 updates](https://github.com/OpenAdaptAI/openadapt-agent/pull/31) (#31) — merged @@ -182,13 +174,19 @@ - [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/openadapt-ml/pull/77) (#77) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/openadapt-ml/pull/75) (#75) — merged - ## openadapt-evals +- [feat(extradup): add the wrong_record fault family and resolve the record by identity](https://github.com/OpenAdaptAI/openadapt-evals/pull/337) (#337) — merged + +- [feat(extradup): add kill-scan CLI for the frozen ExtraDup corpus](https://github.com/OpenAdaptAI/openadapt-evals/pull/336) (#336) — merged + +- [feat(reward): add identity-swap family to the MockMed ExtraDup proof](https://github.com/OpenAdaptAI/openadapt-evals/pull/335) (#335) — merged + +- [docs: freeze M for certified-reward RL Phase-1/pilot](https://github.com/OpenAdaptAI/openadapt-evals/pull/334) (#334) — merged + - [docs(prime): install the environment from the Prime Intellect hub](https://github.com/OpenAdaptAI/openadapt-evals/pull/333) (#333) — merged - [fix(reward): trainer adapters carry the oracle identity the worker requires](https://github.com/OpenAdaptAI/openadapt-evals/pull/332) (#332) — merged @@ -221,14 +219,6 @@ - [fix(evidence): follow the verified-certificate retention receipt](https://github.com/OpenAdaptAI/openadapt-evals/pull/317) (#317) — merged -- [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/openadapt-evals/pull/316) (#316) — merged - -- [fix(windows): require fresh isolation proofs before pool work](https://github.com/OpenAdaptAI/openadapt-evals/pull/315) (#315) — merged - -- [chore(release): prepare 0.94.1](https://github.com/OpenAdaptAI/openadapt-evals/pull/312) (#312) — merged - -- [fix(evidence): match the retention digests the Cloud writer emits](https://github.com/OpenAdaptAI/openadapt-evals/pull/311) (#311) — merged - ## openadapt-capture @@ -287,8 +277,6 @@ - [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/openadapt-privacy/pull/17) (#17) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/openadapt-privacy/pull/16) (#16) — merged - ## openadapt-types @@ -314,8 +302,6 @@ - [chore(policy): sync generated source boundary](https://github.com/OpenAdaptAI/openadapt-types/pull/29) (#29) — merged -- [chore: ignore every .env variant, not one at a time](https://github.com/OpenAdaptAI/openadapt-types/pull/28) (#28) — merged -