Repository navigation
Releases
Releases · fullsend-ai/agents
v0.56.1
Synced with fullsend v0.56.1
What's Changed
- fix(#1999): rewrite skill bash snippets the sandbox hook blocks by @fullsend-ai-coder[bot] in #2000
- test(eval): don't score a review run that only posted a notice by @tpouyer in #2006
- fix(#1500): treat fullsend's withheld notice as a final V2 result by @dhshah13 in #2013
- fix(#1866): use GITHUB_SERVER_URL host in GitHub ops URL handling by @fullsend-ai-coder[bot] in #1873
- fix(#1864): tell triage and review agents their inputs are env vars by @fullsend-ai-coder[bot] in #1870
- fix(#1865): accept triage issue URLs on the run's GitHub host by @fullsend-ai-coder[bot] in #1871
- fix(#1867): restrict triage duplicate targets to open issues by @fullsend-ai-coder[bot] in #1869
- test(#1868): cover the push remote on the configured GitHub host by @fullsend-ai-coder[bot] in #1872
- feat(jira): accept a service account's OAuth 2.0 client credentials by @fullsend-ai-coder[bot] in #2026
- chore(harness): pin fleet images to the v0.56.1-rc.1 release builds by @fullsend-ai-coder[bot] in #2030
New Contributors
Full Changelog: v0.56.0...v0.56.1
v0.56.1-rc.1
Synced with fullsend v0.56.1-rc.1
What's Changed
- fix(#1999): rewrite skill bash snippets the sandbox hook blocks by @fullsend-ai-coder[bot] in #2000
- test(eval): don't score a review run that only posted a notice by @tpouyer in #2006
- fix(#1500): treat fullsend's withheld notice as a final V2 result by @dhshah13 in #2013
- fix(#1866): use GITHUB_SERVER_URL host in GitHub ops URL handling by @fullsend-ai-coder[bot] in #1873
- fix(#1864): tell triage and review agents their inputs are env vars by @fullsend-ai-coder[bot] in #1870
- fix(#1865): accept triage issue URLs on the run's GitHub host by @fullsend-ai-coder[bot] in #1871
- fix(#1867): restrict triage duplicate targets to open issues by @fullsend-ai-coder[bot] in #1869
- test(#1868): cover the push remote on the configured GitHub host by @fullsend-ai-coder[bot] in #1872
New Contributors
Full Changelog: v0.56.0...v0.56.1-rc.1
v0.56.0
Synced with fullsend v0.56.0
What's Changed
- feat(scribe): send explicit no-meeting-notes Slack message on empty s… by @tgeagle081 in #1964
- feat(scribe): include meeting title in Slack notification by @tgeagle081 in #1969
- ci(#1975): move functional, behaviour and vouch workflows off pull_request_target by @fullsend-ai-coder[bot] in #1976
- fix(post-scripts): broaden PEM redaction to cover PGP private-key blocks by @tgeagle081 in #1977
- fix(#409): fetch and check authors before the lease retry in post-code by @fullsend-ai-coder[bot] in #1467
- fix(#1939): have triage inventory issue content before asking by @fullsend-ai-coder[bot] in #1941
- fix(#1989): treat GitLab service accounts as bots in fix cap check by @fullsend-ai-coder[bot] in #1991
- fix(#1983): set signed-push upstream without the fetch refspec by @fullsend-ai-coder[bot] in #1984
- fix(#1995): make the Jira skill snippets parse under the sandbox hook by @fullsend-ai-coder[bot] in #1997
- feat(#1500): dispatch Codex children on collaboration V2 by @dhshah13 in #1971
- feat(#1990)!: read FULLSEND_JIRA_TOKEN in the jira-ro provider by @fullsend-ai-coder[bot] in #1992
- fix(#1848): avoid publishing empty fixture snapshots by @Benkapner in #2008
- chore(harness): pin fleet images to the v0.56.0-rc.1 release builds by @fullsend-ai-coder[bot] in #2021
New Contributors
- @tgeagle081 made their first contribution in #1964
Full Changelog: v0.55.0...v0.56.0
v0.56.0-rc.1
Synced with fullsend v0.56.0-rc.1
What's Changed
- feat(scribe): send explicit no-meeting-notes Slack message on empty s… by @tgeagle081 in #1964
- feat(scribe): include meeting title in Slack notification by @tgeagle081 in #1969
- ci(#1975): move functional, behaviour and vouch workflows off pull_request_target by @fullsend-ai-coder[bot] in #1976
- fix(post-scripts): broaden PEM redaction to cover PGP private-key blocks by @tgeagle081 in #1977
- fix(#409): fetch and check authors before the lease retry in post-code by @fullsend-ai-coder[bot] in #1467
- fix(#1939): have triage inventory issue content before asking by @fullsend-ai-coder[bot] in #1941
- fix(#1989): treat GitLab service accounts as bots in fix cap check by @fullsend-ai-coder[bot] in #1991
- fix(#1983): set signed-push upstream without the fetch refspec by @fullsend-ai-coder[bot] in #1984
- fix(#1995): make the Jira skill snippets parse under the sandbox hook by @fullsend-ai-coder[bot] in #1997
- feat(#1500): dispatch Codex children on collaboration V2 by @dhshah13 in #1971
- feat(#1990)!: read FULLSEND_JIRA_TOKEN in the jira-ro provider by @fullsend-ai-coder[bot] in #1992
New Contributors
- @tgeagle081 made their first contribution in #1964
Full Changelog: v0.55.0...v0.56.0-rc.1
v0.55.0
Synced with fullsend v0.55.0
What's Changed
- fix(#1881)!: migrate agents config to tracker mapping by @ggallen in #1888
- fix(#1881)!: honor tracker allowed targets in retro and triage by @fullsend-ai-coder[bot] in #1889
- fix(#1885): drop duplicate risk-assessment marker from comment by @fullsend-ai-coder[bot] in #1892
- fix(#1795): run scribe content gates before gates that log content by @andrewevansrh in #1880
- test(scripts): isolate HOME in post-fix push-rebase fixtures by @ravelley in #1897
- test(scripts): create Git info directory in public text fixture by @ravelley in #1898
- feat(#519): let the fix agent update stale PR/MR title and body by @fullsend-ai-coder[bot] in #1863
- fix(#1379): bound scribe Slack webhook and drop bc by @fullsend-ai-coder[bot] in #1381
- fix(#1838): accept null summaries for omitted scribe topics by @fullsend-ai-coder[bot] in #1839
- test(#1890): add deterministic state and review action checks by @AshwinUgale in #1891
- fix(#1903): use literal GitLab host in gitlab-forge requests by @fullsend-ai-coder[bot] in #1904
- chore: point CODEOWNERS wildcard to maintainers team by @jwm4 in #1907
- ci(e2e): use expanded dev pool in agents tests by @ralphbean in #1906
- fix(#1909): dispatch the pr-review challenger in the foreground by @fullsend-ai-coder[bot] in #1910
- fix(#1876): send Jira REST calls to JIRA_API_URL, not the site URL by @fullsend-ai-coder[bot] in #1878
- test(post-fix): isolate HOME for the entire test suite by @ravelley in #1905
- fix(scripts): use consistent history reads by @ravelley in #1913
- fix(post-fix): align history traversal with run boundaries by @ravelley in #1915
- feat(#1900)!: support bot and service-account Jira authentication by @fullsend-ai-coder[bot] in #1914
- fix(#1911): skip Jira code runs on human PRs or child issues by @fullsend-ai-coder[bot] in #1912
- fix(#1920): let the code agent download images attached to issues by @fullsend-ai-coder[bot] in #1921
- fix(#1169): write a provisional review result before synthesis by @fullsend-ai-coder[bot] in #1918
- refactor(#1015): port fix harness from forge block to overlays by @fullsend-ai-coder[bot] in #1902
- fix(#1931): separate fix-agent attribution from Next steps list by @fullsend-ai-coder[bot] in #1945
- fix(#1948): use literal GitLab host in review request URLs by @fullsend-ai-coder[bot] in #1949
- chore(harness): pin fleet images to the v0.55.0-rc.1 release builds by @fullsend-ai-coder[bot] in #1956
Full Changelog: v0.54.0...v0.55.0
v0.55.0-rc.1
Synced with fullsend v0.55.0-rc.1
What's Changed
- fix(#1881)!: migrate agents config to tracker mapping by @ggallen in #1888
- fix(#1881)!: honor tracker allowed targets in retro and triage by @fullsend-ai-coder[bot] in #1889
- fix(#1885): drop duplicate risk-assessment marker from comment by @fullsend-ai-coder[bot] in #1892
- fix(#1795): run scribe content gates before gates that log content by @andrewevansrh in #1880
- test(scripts): isolate HOME in post-fix push-rebase fixtures by @ravelley in #1897
- test(scripts): create Git info directory in public text fixture by @ravelley in #1898
- feat(#519): let the fix agent update stale PR/MR title and body by @fullsend-ai-coder[bot] in #1863
- fix(#1379): bound scribe Slack webhook and drop bc by @fullsend-ai-coder[bot] in #1381
- fix(#1838): accept null summaries for omitted scribe topics by @fullsend-ai-coder[bot] in #1839
- test(#1890): add deterministic state and review action checks by @AshwinUgale in #1891
- fix(#1903): use literal GitLab host in gitlab-forge requests by @fullsend-ai-coder[bot] in #1904
- chore: point CODEOWNERS wildcard to maintainers team by @jwm4 in #1907
- ci(e2e): use expanded dev pool in agents tests by @ralphbean in #1906
- fix(#1909): dispatch the pr-review challenger in the foreground by @fullsend-ai-coder[bot] in #1910
- fix(#1876): send Jira REST calls to JIRA_API_URL, not the site URL by @fullsend-ai-coder[bot] in #1878
- test(post-fix): isolate HOME for the entire test suite by @ravelley in #1905
- fix(scripts): use consistent history reads by @ravelley in #1913
- fix(post-fix): align history traversal with run boundaries by @ravelley in #1915
- feat(#1900)!: support bot and service-account Jira authentication by @fullsend-ai-coder[bot] in #1914
- fix(#1911): skip Jira code runs on human PRs or child issues by @fullsend-ai-coder[bot] in #1912
- fix(#1920): let the code agent download images attached to issues by @fullsend-ai-coder[bot] in #1921
- fix(#1169): write a provisional review result before synthesis by @fullsend-ai-coder[bot] in #1918
- refactor(#1015): port fix harness from forge block to overlays by @fullsend-ai-coder[bot] in #1902
- fix(#1931): separate fix-agent attribution from Next steps list by @fullsend-ai-coder[bot] in #1945
- fix(#1948): use literal GitLab host in review request URLs by @fullsend-ai-coder[bot] in #1949
Full Changelog: v0.54.0...v0.55.0-rc.1
v0.54.0
Synced with fullsend v0.54.0
What's Changed
- feat(#1857)!: select pr-review personas from frontmatter, drop REVIEW_UX_ENABLED by @fullsend-ai-coder[bot] in #1858
- test(#1859): bump fullsend and run the custom-persona review scenario by @fullsend-ai-coder[bot] in #1860
- fix(#1701): preserve first-attempt artifacts when retrying cases by @fullsend-ai-coder[bot] in #1751
- fix(#1567): key projection suppression off failed sub-agent name by @fullsend-ai-coder[bot] in #1705
- fix(#1853): tighten the Codex wave after maintainer review by @dhshah13 in #1856
- fix(#1818): keep finding ids on their stamped review threads by @fullsend-ai-coder[bot] in #1819
- refactor(#1861): keep each eval attempt in attempts/ with final.json by @fullsend-ai-coder[bot] in #1875
- feat(#1271): scale code-agent discovery to precisely scoped issues by @fullsend-ai-coder[bot] in #1874
- fix(#1833): compare GitHub owner logins case-insensitively in PR recovery by @fullsend-ai-coder[bot] in #1836
- feat(#1654): scope review to own commits on stacked/cherry-picked PRs by @fullsend-ai-coder[bot] in #1661
- fix(scripts): support BSD sed by @ravelley in #1840
- chore(harness): pin fleet images to the v0.54.0-rc.1 release builds by @fullsend-ai-coder[bot] in #1884
New Contributors
Full Changelog: v0.53.0...v0.54.0
v0.54.0-rc.1
Synced with fullsend v0.54.0-rc.1
What's Changed
- feat(#1857)!: select pr-review personas from frontmatter, drop REVIEW_UX_ENABLED by @fullsend-ai-coder[bot] in #1858
- test(#1859): bump fullsend and run the custom-persona review scenario by @fullsend-ai-coder[bot] in #1860
- fix(#1701): preserve first-attempt artifacts when retrying cases by @fullsend-ai-coder[bot] in #1751
- fix(#1567): key projection suppression off failed sub-agent name by @fullsend-ai-coder[bot] in #1705
- fix(#1853): tighten the Codex wave after maintainer review by @dhshah13 in #1856
- fix(#1818): keep finding ids on their stamped review threads by @fullsend-ai-coder[bot] in #1819
- refactor(#1861): keep each eval attempt in attempts/ with final.json by @fullsend-ai-coder[bot] in #1875
- feat(#1271): scale code-agent discovery to precisely scoped issues by @fullsend-ai-coder[bot] in #1874
- fix(#1833): compare GitHub owner logins case-insensitively in PR recovery by @fullsend-ai-coder[bot] in #1836
- feat(#1654): scope review to own commits on stacked/cherry-picked PRs by @fullsend-ai-coder[bot] in #1661
- fix(scripts): support BSD sed by @ravelley in #1840
New Contributors
Full Changelog: v0.53.0...v0.54.0-rc.1
v0.53.0
Synced with fullsend v0.53.0
What's Changed
- test(code): exercise the real push-retry instead of mirrored logic by @Benkapner in #1834
- fix(code): recover the PR when create succeeds but the response is lost by @Benkapner in #1832
- feat(#1067): add CEL dispatch triggers to code and fix harnesses by @fullsend-ai-coder[bot] in #1842
- feat(#1800): disclose findings hidden by the severity threshold by @Benkapner in #1827
- fix(#1801): scan only added lines in post-fix merge secret checks by @fullsend-ai-coder[bot] in #1802
- feat(#1841): write triage and code results to the originating Jira issue by @fullsend-ai-coder[bot] in #1843
- fix(#1824): attribute review findings to the evaluated run by @fullsend-ai-coder[bot] in #1825
- fix(#1845): scan commit identity in post-fix secret check by @fullsend-ai-coder[bot] in #1847
- feat(#1500): dispatch sub-agents in waves on Codex by @dhshah13 in #1849
- docs(agents): align AGENTS.md with COMMITS.md and CONTRIBUTING.md SSOT by @durandom in #1749
- chore(harness): pin fleet images to the v0.53.0-rc.1 release builds by @fullsend-ai-coder[bot] in #1855
New Contributors
Full Changelog: v0.52.0...v0.53.0
v0.53.0-rc.1
Synced with fullsend v0.53.0-rc.1
What's Changed
- test(code): exercise the real push-retry instead of mirrored logic by @Benkapner in #1834
- fix(code): recover the PR when create succeeds but the response is lost by @Benkapner in #1832
- feat(#1067): add CEL dispatch triggers to code and fix harnesses by @fullsend-ai-coder[bot] in #1842
- feat(#1800): disclose findings hidden by the severity threshold by @Benkapner in #1827
- fix(#1801): scan only added lines in post-fix merge secret checks by @fullsend-ai-coder[bot] in #1802
- feat(#1841): write triage and code results to the originating Jira issue by @fullsend-ai-coder[bot] in #1843
- fix(#1824): attribute review findings to the evaluated run by @fullsend-ai-coder[bot] in #1825
- fix(#1845): scan commit identity in post-fix secret check by @fullsend-ai-coder[bot] in #1847
- feat(#1500): dispatch sub-agents in waves on Codex by @dhshah13 in #1849
- docs(agents): align AGENTS.md with COMMITS.md and CONTRIBUTING.md SSOT by @durandom in #1749
New Contributors
Full Changelog: v0.52.0...v0.53.0-rc.1