• v2.29.0 aa276ecf19

    agent-compose v2.29.0
    All checks were successful
    release / release (push) Successful in 1m22s
    Stable

    coilysiren released this 2026-08-17 18:17:17 +00:00 | 34 commits to main since this release

    • Merge pull request 'feat(color): derive role favorites from the whole roster so they stay tellable apart' (#300) from aos/claude/favorite-color-spread into main (aa276ec)
    • feat(color): derive role favorites from the whole roster so they stay tellable apart (105c04b)
    Downloads
  • v2.28.0 81e317cc00

    agent-compose v2.28.0
    All checks were successful
    release / release (push) Successful in 1m22s
    Stable

    coilysiren released this 2026-08-17 17:56:32 +00:00 | 36 commits to main since this release

    • Merge pull request 'feat(person): give exec and ai a personality each owns, closing the last emphasis collision' (#299) from aos/claude/personality-catalog-gaps into main (81e317c)
    • feat(person): give exec and ai a personality each owns, closing the last emphasis collision (45fd7f5)
    Downloads
  • v2.27.0 b13d2a0df7

    agent-compose v2.27.0
    All checks were successful
    release / release (push) Successful in 1m36s
    Stable

    coilysiren released this 2026-08-17 17:55:14 +00:00 | 38 commits to main since this release

    • Merge pull request 'feat(schema): let a role narrow the skills it receives from a mounted provider (#297)' (#298) from aos/claude/gh297-binding-skill-selector into main (b13d2a0)
    • feat(schema): let a role narrow the skills it receives from a mounted provider (#297) (ab46ebe)
    Downloads
  • v2.26.0 caa89e0010

    agent-compose v2.26.0
    All checks were successful
    release / release (push) Successful in 1m16s
    Stable

    coilysiren released this 2026-08-17 17:21:41 +00:00 | 40 commits to main since this release

    • Merge pull request 'refactor(person): break the shared personality template so the sixteen stop converging' (#296) from aos/claude/personality-prose-distinctiveness into main (caa89e0)
    • refactor(person): break the shared personality template so the sixteen stop converging (1be1827)
    • Merge pull request 'refactor(evalkit): adopt aos-eval, and delete the half it now owns (#1104)' (#294) from aos/claude/gh56-adopt-aos-eval into main (8a1cf25)
    • Merge origin/main into the aos-eval adoption (#295) (5210416)
    • fix(evalkit): pin the aos-eval tag that exists, and lock against it (#295) (43ae7c1)
    • refactor(evalkit): adopt aos-eval, and delete the half it now owns (#1104) (23488b4)
    Downloads
  • v2.25.0 d2c6f730af

    agent-compose v2.25.0
    All checks were successful
    release / release (push) Successful in 1m32s
    Stable

    coilyco-ops released this 2026-08-17 05:08:47 +00:00 | 46 commits to main since this release

    • docs: merge 65 docs to 40 and advance the hook pin (d2c6f73)
    • chore(docs): declare the large documentation band (b994635)
    Downloads
  • v2.24.0 979821d009

    agent-compose v2.24.0
    All checks were successful
    release / release (push) Successful in 1m11s
    Stable

    coilysiren released this 2026-08-16 05:52:16 +00:00 | 48 commits to main since this release

    • Merge pull request 'refactor(just): retire per-repo ward exec, carry every verb in the justfile' (#291) from aos/claude/xu59-just into main (979821d)
    • refactor(just): retire per-repo ward exec, carry every verb in the justfile (0cfe8ea)
    • Merge pull request 'feat(evalkit): project a committed run into a display payload, one way' (#290) from aos/claude/xu59-eval-export into main (56e7284)
    • feat(evalkit): project a committed run into a display payload, one way (755b963)
    Downloads
  • v2.23.0 15ce125dd5

    agent-compose v2.23.0
    All checks were successful
    release / release (push) Successful in 1m12s
    Stable

    coilysiren released this 2026-08-16 04:22:05 +00:00 | 52 commits to main since this release

    • Merge pull request 'fix(launch): bring two comment blocks back under the cap' (#289) from aos/claude/xu59-comment-caps into main (15ce125)
    • fix(launch): bring two comment blocks back under the cap (4498ff8)
    • Merge pull request 'ci: validate a pull request before it can become a red main' (#288) from aos/claude/xu59-pr-ci into main (17dfbac)
    • docs(features): drop the CI entry, which the rule excludes (4d68009)
    • ci: validate a pull request before it can become a red main (c0877c7)
    • Merge pull request 'feat(launch): export the composed role for commit attribution' (#287) from aos/claude/xu59-role-attribution into main (3d1bc50)
    • feat(launch): export the composed role for commit attribution (51a7d4b)
    • docs(readme): match the lead line to the description it states (9039e92)
    • feat(banner): remint the banner on the repository description (856c228)
    • Update README.md (6118b21)
    Downloads
  • v2.22.0 317e5f0bb3

    agent-compose v2.22.0
    All checks were successful
    release / release (push) Successful in 1m42s
    Stable

    coilyco-ops released this 2026-08-14 06:49:19 +00:00 | 62 commits to main since this release

    • feat: add acompose whoami, the one answer to what a session is called (317e5f0)
    Downloads
  • v2.21.0 839a118ee0

    agent-compose v2.21.0
    All checks were successful
    release / release (push) Successful in 1m44s
    Stable

    coilyco-ops released this 2026-08-14 06:25:07 +00:00 | 63 commits to main since this release

    • feat: name the running agent with its dictatable short id (839a118)
    • Update .ward/ward.yaml (65843bb)
    • Merge pull request 'feat(agents): declare the git workflow in AGENTS.md frontmatter' (#282) from aos/claude/mg96-fm into main (e5d6376)
    • feat(agents): declare the git workflow in AGENTS.md frontmatter (c1dd4db)
    Downloads
  • v2.20.0 85352838c6

    agent-compose v2.20.0
    All checks were successful
    release / release (push) Successful in 1m18s
    Stable

    coilyco-ops released this 2026-08-14 04:16:05 +00:00 | 67 commits to main since this release

    • fix(ci): stop calling the scorecard command #262 deleted (8535283)
    • feat(schema): let a request name the seat it composes (71b112b)
    • chore: refresh the managed agentic-os hook block to v0.18.0 (d5783b7)
    • Merge pull request 'refactor!: retire the driver-plus-reviewer eval stack (#262)' (#269) from evals/262-board-run into main (b486f88)
    • Merge pull request 'docs(agents): state the temp-clone expectation' (#268) from claude/agents-temp-clone-note into main (0445e5a)
    • docs(agents): state the temp-clone expectation (8fedc3e)
    • refactor!: retire the driver-plus-reviewer eval stack (#262) (b3be43f)
    • fix(assets): quiet the banner for the page it sits on (017c5d2)
    • Update README.md (a19ae62)
    • docs: lead the README with the banner (e8444c8)
    • fix(evalkit): scope the annotation count and unbreak long critiques (#262) (8ea6681)
    • docs(agents): state the temp-clone expectation (a510818)
    • feat(assets): ship the agent-compose banner (ca223ad)
    • refactor(evalkit)!: delete the regex discriminator tier (#262) (f13dea3)
    • test(evals): probe absent-authority before editing doctrine (#262) (fa02b09)
    • test(evals): preserve the graded ops pilot board (#262) (41ab299)
    • fix(evalkit): give the annotate verb its paths (#262) (963b4d3)
    • fix(evalkit): report item analysis instead of gating on it (#262) (736a327)
    • refactor(evaluation)!: test one commodity subject tier, retire the lane axis (0c94735)
    • test(evals): make the ops crossings incidental and the patterns artifact-shaped (#262) (7e8532a)
    • fix(evalkit): resolve run paths before Inspect changes directory (#262) (3ddee86)
    • test(evals): author the ops pilot slice of the board (#262) (48ffbe9)
    • Update README.md (0dd9fcc)
    • feat(assets): add the agent-compose mark (13bea51)
    • Update README.md (df3e870)
    • Merge pull request 'evalkit: run through Inspect, pydantic wire types, pattern discriminators' (#266) from adjacency/262-role-adjacents into main (301816c)
    • feat(evalkit): make discriminators machine-checkable patterns (e6bbfb9)
    • feat(evalkit): run through Inspect and model the wire types with pydantic (dc426ce)
    • Merge pull request 'refactor(evalkit): adopt the reference vocabulary and add axial coding' (#265) from adjacency/262-role-adjacents into main (4035f10)
    • refactor(evalkit): adopt the reference vocabulary and add axial coding (84ede51)
    • Merge pull request 'feat(evalkit): wire the transport and the composed-prompt producer' (#264) from adjacency/262-role-adjacents into main (0c8b733)
    • feat(evalkit): wire the transport and the composed-prompt producer (8a7629d)
    • Merge pull request 'Role adjacency in KDL, and evalkit as the Python half of the eval system' (#263) from adjacency/262-role-adjacents into main (922469f)
    • docs: add reference pages for the six external eval reference points (3aa9faf)
    • feat(evalkit): group grading by role and print the charter once per group (533dbe8)
    • feat(evalkit): state what a within-role case tests, not whose work it is (54c1df8)
    • feat(evalkit): drop the composed-meld personality case (45e9bfe)
    • feat(evalkit): interpolate roster context into boundary case descriptors (5dd7082)
    • feat(evalkit): derive the eval case list from the roster (470fab2)
    • feat(evalkit): add the Python orchestration half of the eval system (9d22c27)
    • feat(person): declare role adjacency in KDL with a fixed out-degree (cdd293c)
    • Merge pull request 'chore(hooks): pull the aos pre-commit suite to aos-precommit-v0.16.0' (#261) from chore/precommit-bump into main (67fd5bc)
    • chore(hooks): pull the aos pre-commit suite to aos-precommit-v0.16.0 (eb0350a)
    • Merge pull request 'refactor(person): flat data layout, freeform personalities, bounded prose (#233 #251 #255 #256 #257 #258)' (#260) from prose/257-258-233-second-person into main (8341c56)
    • feat(person): bound every shipped prose entry at both ends (#256) (9f8730d)
    • refactor(person): one flat directory per entity under data/ (#255) (43739fd)
    • refactor(person): make personality bodies freeform prose (#251) (09c9ec0)
    • refactor(prose): second person, no assumed human, identity-led intro (#233, #257, #258) (67518b7)
    • Merge pull request 'refactor(person)!: rename meld to boundary and name behaviors (#254)' (#259) from boundary/254-primitive-rename into main (7362c58)
    • feat(person)!: name boundaries for the behavior they remove (#254) (fb94739)
    • refactor(person)!: rename the meld primitive to boundary (#254) (d09e97d)
    • Merge pull request '#253' from comms/rescope-counterpart into main (4ce2537)
    • fix(resolver): let a source bind skills to roles the roster lacks (2f4fbba)
    • Merge pull request 'fix: give a role launch its own user-scoped load points' (#252) from native-home/968-user-scope-projection into main (7b92339)
    • refactor(person): scope comms to socially toned surfaces, add counterparts (3534668)
    • fix(nativelaunch): project the role into the session home, base included (f1097ac)
    • Merge pull request 'feat(person): make evidence acquisition an active shared obligation (#248)' (#249) from meld/248-evidence-acquisition into main (8846e78)
    • fix(cascade): wire claude's global skill load point by default (5b21ac8)
    • refactor(person): scope the evidence meld to three roles (#248) (8eb92b6)
    • chore(ward): declare the merge-remote-main landing policy (a1b93ae)
    • feat(person): make evidence acquisition an active shared obligation (#248) (5823c9c)
    • refactor(roster)!: rename the strats role to exec, Executive Strategist (e5c3b93)
    • refactor(palette): carry the // separator into the web explorer (02b0590)
    • refactor(render): replace the middle-dot separator with // (a09c1f6)
    • feat(identity): annotate every terminal surface with pronouns and role (cd0fdff)
    • merge: reconcile the reviewer with main's arm lane (#231) (a96b5a4)
    • Merge pull request 'evals/240-baseline' (#247) from evals/240-baseline into main (9764756)
    • fix(evals): seal driver sessions from real systems (#240) (c862e46)
    • docs: bring the size caps back under the limit (#240) (2f52f55)
    • test(evals): drive the Claude frontier lane with global memory relocated (#240) (8994cb6)
    • Merge pull request 'test(evals): compare opus-5 high against sonnet-5 medium (#240)' (#245) from evals/240-model-arms into main (e084345)
    • test(evals): re-earn a baseline end to end (#231) (24eb173)
    • test(evals): compare opus-5 high against sonnet-5 medium (#240) (1b90504)
    • docs(features): fit the merged inventory under the size cap (#231) (cba79ac)
    • Merge pull request 'feat(person): add shared role melds for cross-role doctrine (#231)' (#243) from meld/231-shared-role-melds into main (53be537)
    • Merge remote-tracking branch 'origin/main' into meld/231-shared-role-melds (5dd6565)
    • Merge pull request 'test(evals): add an executable evaluation driver (#240)' (#244) from evals/240-model-arms into main (1e5c409)
    • test(evals): add an executable evaluation driver (#240) (e28f927)
    • feat(person): add shared role melds for cross-role doctrine (#231) (6b22254)
    Downloads