Skip to content

Record guarded agent autonomy acceptance #95

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Techris93 merged 1 commit into from
Jul 27, 2026
Merged

Record guarded agent autonomy acceptance #95

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion docs/implementation-checkpoints.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@

## 108 Guarded Agent Research Completion

Status: implementation complete; full verification and Mission Control release are pending.
Status: complete in the local pilot and merged.

Added an optional `agent_review` mode for the Local Codex/OpenCodex bridge. When all bounded static-analysis jobs finish, Core accepts pipeline-scoped proposals, records an evidence-linked verdict, and reruns publication safety without requiring card-by-card operator acceptance. Verdict output is schema constrained and then independently guarded by Core: low-confidence work remains inconclusive, a credible verdict requires advisory-backed or sandbox evidence, unresolved contradictions and unsupported claims reduce certainty, and absence from a local repository can never establish package benignness or make ecosystem intelligence non-actionable.

Expand All @@ -20,6 +20,10 @@ Follow-up: bridge service installation now persists the selected OpenCodex provi

Live acceptance exposed a retracted-package recovery gap: an exact NuGet version previously collected and hashed had disappeared from the registry. Investigation pipelines now prefer a fresh official-registry collection but may reuse an exact prior quarantine artifact after independently verifying ecosystem, package, version, byte size, SHA-256, regular-file state, and configured size limit. Missing or altered cache entries still fail closed. The recorded step makes registry unavailability and quarantine reuse explicit, and no package code is executed.

Live acceptance: pipeline `RPL-37985992E2512C07` recovered `nuget:Braintree.Net@3.35.7` after the registry returned HTTP 404, reused the exact hash-verified quarantine artifact, freshly collected the legitimate reference package, and completed all three structured analyses through `google-antigravity/gemini-3.6-flash`. Guarded agent review accepted 18 revision-scoped proposals and recorded a `likely` verdict at 85% confidence. Publication preflight remained `needs_approval`; sandbox submission, disclosure delivery, and publication all remained false. The earlier Kimi/Grok provider-credit failures remain in the job audit history.

Release acceptance: Core PRs #92, #93, and #94 and Mission Control PRs #32 and #33 passed their Python-version matrix, MCP, security, dependency, Cloudflare, dashboard, and browser checks before merge. The local launchd bridge is running in `agent_review` mode with Gemini 3.6 persisted in its fixed command and no provider credentials in the service definition.

## 107 Hermes Agent v1.0.0 Integration

Status: complete in production.
Expand Down
2 changes: 1 addition & 1 deletion docs/research-automation.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ Install or update the background bridge in agent-review mode:
```bash
secopsai intelligence bridge service install \
--autonomy-mode agent_review \
--model kimi/kimi-k2.7-code-highspeed
--model google-antigravity/gemini-3.6-flash
```

Complete an already-waiting pipeline from the CLI:
Expand Down
Loading