v0.16.0

Open

No due date

•Last updated Sep 14, 2026

Release v0.16.0: the tail. Entry: the last consolidation defects, and capabilities whose blocker sits in an overflow version. Never ahead of an open blocker.

0% complete

feat(llc): project-basis reporting — the dotted line over the org chart
mrveiss/AutoBot-AI#15766 · opened on Sep 5, 2026 · v0.16.0
One hook parses Python to avoid string-literal false positives; the print guard greps and cannot
mrveiss/AutoBot-AI#16011 · opened on Sep 8, 2026 · v0.16.0
2
code-sync in the maintenance GUI never reaches a terminal state
mrveiss/AutoBot-AI#16021 · opened on Sep 8, 2026 · v0.16.0
5
/slm/setup cannot get past provision
mrveiss/AutoBot-AI#16022 · opened on Sep 8, 2026 · v0.16.0
A node that is online-but-broken cannot be repaired: enrollment gates on agent reachability, not service health
mrveiss/AutoBot-AI#16078 · opened on Sep 8, 2026 · v0.16.0
bug(hooks): a hook emits permissionDecision "warn", which is not in the contract and does nothing (#15956 family)
mrveiss/AutoBot-AI#16079 · opened on Sep 8, 2026 · v0.16.0
1
Three pushes produced no synchronize event and nothing detected it: a PR with zero check-runs reads as queued
mrveiss/AutoBot-AI#16091 · opened on Sep 8, 2026 · v0.16.0
1
bug(guards): git-scope.sh runs every git call unscrubbed, and its header cites the scrubbing issue (#15506 family)
mrveiss/AutoBot-AI#16109 · opened on Sep 9, 2026 · v0.16.0
1
feat(rlm): reflection is a single scalar — no itemized checks and no re-audit after repair
mrveiss/AutoBot-AI#16112 · opened on Sep 9, 2026 · v0.16.0
eval: the trajectory-eval gate is non-blocking with 3 goldens — broaden the corpus so review can safely shorten
mrveiss/AutoBot-AI#16115 · opened on Sep 9, 2026 · v0.16.0
1
Worktree ceiling denies every session and its documented override cannot work
mrveiss/AutoBot-AI#16152 · opened on Sep 9, 2026 · v0.16.0
2
Eight sites rely on inline codeql/lgtm suppressions that this repo documents as non-functional; two alerts are open behind them
mrveiss/AutoBot-AI#16172 · opened on Sep 10, 2026 · v0.16.0
4
A replay timeout is reported as a regression, which is a verdict about a candidate that was never scored
mrveiss/AutoBot-AI#16192 · opened on Sep 10, 2026 · v0.16.0
ci: a conflicting PR's merge ref freezes at its last good merge while base.sha moves on — 290 commits stale, invisible from the ref
mrveiss/AutoBot-AI#16216 · opened on Sep 10, 2026 · v0.16.0
3
pricing(subscriptions): per-account monthly subscription plans alongside per-token billing
mrveiss/AutoBot-AI#16232 · opened on Sep 10, 2026 · v0.16.0
pricing(guard): fail the build on any hardcoded model price, and on manufactured freshness
mrveiss/AutoBot-AI#16233 · opened on Sep 10, 2026 · v0.16.0
1
fix(monitoring): Grafana dashboards reference a Prometheus datasource uid that no provisioning defines
mrveiss/AutoBot-AI#16290 · opened on Sep 11, 2026 · v0.16.0
company-os: fold business_intelligence_dashboard's cost and ROI analysis into Company OS, and system-wide stats into the SLM
mrveiss/AutoBot-AI#16307 · opened on Sep 11, 2026 · v0.16.0
1
pricing(local-models): a local/self-hosted model now raises UnpricedModel instead of accruing $0
mrveiss/AutoBot-AI#16316 · opened on Sep 11, 2026 · v0.16.0
3
site.yml's Frontend play references a nonexistent role frontend_app
mrveiss/AutoBot-AI#16342 · opened on Sep 11, 2026 · v0.16.0
fix(backend): the Redis service manager queries the SLM with an empty node id and no credential, so status is always unknown
mrveiss/AutoBot-AI#16346 · opened on Sep 11, 2026 · v0.16.0
bug(settings): require_settings_admin records a username-less admin session as the internal service
mrveiss/AutoBot-AI#16468 · opened on Sep 12, 2026 · v0.16.0
chore(scripts): test_first_remediation.py still bases new worktrees on origin/Dev_new_gui
mrveiss/AutoBot-AI#16517 · opened on Sep 12, 2026 · v0.16.0
2
Per-provider concurrency cap does not cover stream_completion() — only chat_completion() is bounded
mrveiss/AutoBot-AI#16538 · opened on Sep 12, 2026 · v0.16.0
1

Navigation Menu

v0.16.0

feat(llc): project-basis reporting — the dotted line over the org chart

One hook parses Python to avoid string-literal false positives; the print guard greps and cannot

code-sync in the maintenance GUI never reaches a terminal state

/slm/setup cannot get past provision

A node that is online-but-broken cannot be repaired: enrollment gates on agent reachability, not service health

bug(hooks): a hook emits permissionDecision "warn", which is not in the contract and does nothing (#15956 family)

Three pushes produced no synchronize event and nothing detected it: a PR with zero check-runs reads as queued

bug(guards): git-scope.sh runs every git call unscrubbed, and its header cites the scrubbing issue (#15506 family)

feat(rlm): reflection is a single scalar — no itemized checks and no re-audit after repair

eval: the trajectory-eval gate is non-blocking with 3 goldens — broaden the corpus so review can safely shorten

Worktree ceiling denies every session and its documented override cannot work

Eight sites rely on inline codeql/lgtm suppressions that this repo documents as non-functional; two alerts are open behind them

A replay timeout is reported as a regression, which is a verdict about a candidate that was never scored

ci: a conflicting PR's merge ref freezes at its last good merge while base.sha moves on — 290 commits stale, invisible from the ref

pricing(subscriptions): per-account monthly subscription plans alongside per-token billing

pricing(guard): fail the build on any hardcoded model price, and on manufactured freshness

fix(monitoring): Grafana dashboards reference a Prometheus datasource uid that no provisioning defines

company-os: fold business_intelligence_dashboard's cost and ROI analysis into Company OS, and system-wide stats into the SLM

pricing(local-models): a local/self-hosted model now raises UnpricedModel instead of accruing $0

site.yml's Frontend play references a nonexistent role `frontend_app`

fix(backend): the Redis service manager queries the SLM with an empty node id and no credential, so status is always unknown

bug(settings): require_settings_admin records a username-less admin session as the internal service

chore(scripts): test_first_remediation.py still bases new worktrees on origin/Dev_new_gui

Per-provider concurrency cap does not cover stream_completion() — only chat_completion() is bounded

Issues