Daily Fro Bot Report — 2026-10-13 (UTC) #3990

Daily Fro Bot Report — 2026-10-13 (UTC)

Snapshot starts 2026-10-13T06:30:27.125478+00:00. Categories 1–4 are attributed to the separate remediation readback; this pass ran categories 5–8. No repair PR was produced.

Run Summary

✅ verified-clean · ⚠️/❌ warning/error · ❔ could-not-check. Empty, unavailable, stale or denied evidence is never an all-clear.

Errored PRs

None. Remediation inspected check runs and separate legacy statuses for fro-bot/.github/pull/3989, fro-bot/.github/pull/3988, fro-bot/.github/pull/3987, fro-bot/.github/pull/3984. Current readbacks retain successful completed checks and legacy private-leak statuses; no repair comments or PRs were produced. The semantic review failure on #3989 is reported under Progressive Improvement.

Security

Remediation evidence: three open transitive development Dependabot alerts (two medium, one low); live npm audit three moderate/one low, zero high/critical; override-floor check passes. All 17 external Action/reusable-workflow packages were checked against the GitHub Advisory Database, with no affected pin. fro-bot/.github/security/dependabot/59, fro-bot/.github/security/dependabot/77, fro-bot/.github/security/dependabot/79. Next: compatible Renovate updates through fro-bot/.github/issues/2828. Scorecard’s high rule label is not the package-advisory severity. Sources: npm audit registry, GitHub Dependabot/Advisory Database, upstream tag resolution; cross-major affected ranges included, latest-major drift not surveyed by remediation. No security repair PR.

Control-Plane Integrity

Remediation evidence: 31 workflows/two composites parse; 146 SHA/version references verified against 17 upstream tags; 74 production scripts load under Node 24 strip-only. Explicit App scopes, same-repository data authority, privacy gates and live protections verified. Surplus CodeQL reads remain. Effective execution guard is unavailable/unverified, tracked with fro-bot/.github/issues/3979. No repair PR.

Code Quality

Remediation evidence: pnpm bootstrap, pnpm check-types, pnpm lint, pnpm test pass; 102 test files / 5,358 passed / three todo. This pass separately passed pnpm lint, git diff --check, and read-only wiki validation: 47 pages, zero deterministic findings on the edited topic/index. Two existing broken LICENSE links remain under fro-bot/.github/issues/3903; four unrelated advisory findings. Install/build-writing commands were not repeated under the non-mutating-shell/report-only contract.

Oversight

Paginated authenticated /user/repos and /user/orgs, followed by repository listings for every returned organization (bfra-me, psware-ps2, pro-actions), succeeded. “Can see” means returned records with read/pull permission. The psware-ps2 listing is empty, so that organization supplies no repository-health evidence. Only confirmed-public identities/findings appear below. Activity endpoints all succeeded.

Public snapshot: 225 open issues, 138 open PRs, 14 new issues in 24h, 82 issues inactive >30d, 94 PRs aged >7d, 56 PRs inactive >14d, 6 unassigned bugs. PR age uses creation; inactivity uses the latest update, commit, review, comment or state-change activity, read with pagination. Snapshot counts precede report creation/retirement.

Top three hotspots — ranked by distinct qualifying URLs: new/stale issues, aged/stale PRs, unassigned bugs, accessible alerts and latest failing default-head contexts; overlaps count once, alerts are not unique exploitable root causes.

• marcusrbrown/extend-vscode: 98 — marcusrbrown/extend-vscode/security/dependabot/211. Next: prioritize existing dedicated security work, then review inactive queues; preserve package/image advisory distinctions.

• marcusrbrown/sparkle: 64 — marcusrbrown/sparkle/security/dependabot/320. Next: prioritize existing dedicated security work, then review inactive queues; preserve package/image advisory distinctions.

• marcusrbrown/containers: 61 — marcusrbrown/containers/security/dependabot/51. Next: prioritize existing dedicated security work, then review inactive queues; preserve package/image advisory distinctions.

New issues — next: triage each linked issue: bfra-me/renovate-action/issues/3927, bfra-me/works/issues/5108, fro-bot/agent/issues/1817, fro-bot/agent/issues/1805, fro-bot/space-bus/issues/209, marcusrbrown/.dotfiles/issues/2858, marcusrbrown/dev-like/issues/164, marcusrbrown/infra/issues/1527, marcusrbrown/marcusrbrown.github.io/issues/465, marcusrbrown/mothership/issues/145, marcusrbrown/systematic/issues/1106, marcusrbrown/systematic/issues/1105, marcusrbrown/systematic/issues/1099, marcusrbrown/tokentoilet/issues/1612

Unassigned bugs — next: reproduce and assign each: bfra-me/ha-addon-repository/issues/569, marcusrbrown/infra/issues/1521, marcusrbrown/marcusrbrown.com/issues/517, marcusrbrown/marcusrbrown.com/issues/465, marcusrbrown/systematic/issues/1099, marcusrbrown/systematic/issues/740

Failing current default-head contexts — next: inspect failed steps/current inputs before retrying: marcusrbrown/extend-vscode/actions/runs/38191288729/job/114624417018, marcusrbrown/marcusrbrown.github.io/actions/runs/38212153327/job/114689530095, marcusrbrown/marcusrbrown.github.io/actions/runs/38212153327/job/114688511911, marcusrbrown/marcusrbrown.github.io/actions/runs/38212153327/job/114688511657

Latest check per app/name and latest legacy status per context were inspected separately. Portfolio failures are the latest check executions on its current head; no new execution was triggered here. Empty check-run and legacy-status histories — next: confirm intended CI/archive coverage: fro-bot/tokentoilet, marcusrbrown/encode-2024-q3-grp14-jobseeker-ai, marcusrbrown/gala-chain-code, pro-actions/peter-murray_workflow-application-token-action.

Dependabot: 249 accessible public alerts; 6/40 endpoints unavailable, with status/repository links above. Accessible empty lists are not proof a scanner ran.

Code Scanning: 189 accessible public alerts; 25/40 endpoints unavailable, with status/repository links above. Accessible empty lists are not proof a scanner ran.

secret-scanning: 0 accessible public alerts; 33/40 endpoints unavailable, with status/repository links above. Accessible empty lists are not proof a scanner ran.

Gateway rollout awareness — category 8

#3512 remains Open / Project 1 In Progress, waiting; incomplete acceptance is consistent. All 29 Project items were read. The dedicated tracker comment acknowledges closures, but the issue body still says Open:

Dashboard #585 and #586: body Open / Project Todo, consistent. Checklist items still owe authenticated production acceptance; Done/released/deployed is not behavioral verification.

• Body live contract 1.8.0 versus fresh health 1.9.0.

• Body gateway v0.118.2 versus infra pin v0.119.2 and successful actual deploy job, completed October 13 04:04:07 UTC, logs referencing v0.119.2.

• Body dashboard 2026.10.15 versus infra ledger 2026.10.21 versus successful actual deploy job 2026.10.23, completed 04:02:23 UTC. Its post-convergence evidence names 2026.10.23@sha256:09271b4083632e01135482b8a4292903a604fe7fe6a4cf28b9f7f212993a23e1.

• Upstream latest-release truth: agent v0.122.0 and dashboard 2026.10.23; tagged contract files declare 1.10.0 / 1.9.0. Released-contract skew remains; public live health is 1.9.0. Minor contract/release drift included, no separate cross-major gateway survey.

Next: dedicated tracker reconciles body/ledger claims and verifies paired compatibility before a newer gateway cutover. Authenticated SSE ready, question answer/skip/settlement, checkout projection, cancellation, expired-output recovery, logout/CSRF and privacy acceptance were not checked here. No tracker comment or Project edit.

Cross-Project Intelligence

Every tracking row was considered. Public source inspection fetched 31 tracked trees / 220 workflow files, plus available manifests/agent guidance; no public tree truncation or fetch error. Coverage is partial: fro-bot/fro-bot.github.io, fro-bot/systematic have empty default workflow trees. This is configured-source inspection, not execution verification. Adoptable findings only:

1. Agent harness integration: explicit working-dir / response-none candidate delivery, trusted finalizer, single-job OIDC authority, pre-agent no-post App mint, step-scoped key/token exposure and expiry fail-closed. Adopt these boundaries when extending ingestion.

2. Systematic diagnostics: !cancelled() plus actual suite outcome preserves diagnostics after a failed suite without converting failure into success. Adopt for handoff/producer diagnosis.

3. Infra release alert: owner-only synthetic validation, paginated marker/title/label identity, exact readback with bounded retries and duplicate fail-closed. Adopt those validation-mode checks; its normal path still has a bounded lookup, so do not copy it as proof of complete production dedup.

No implementation changes in this category.

Progressive Improvement

Version authority: individually queried npm registry latest stable tags; major drift included. Manifest:

All 31 committed workflows are active. Current-main Main run passes lint/types/tests, strip-only loading, workflow and dist checks; PR-only authority/mutation jobs correctly skip on push. Missing preset-validation gate remains #3793. No stale production TODO/FIXME found in scripts/workflows: one negative-test fixture and this scan’s prompt text. Shared setup/pnpm conventions match documented workflow exceptions; preflight/guard uncertainty remains #3979.

Learning proposals: one open, zero >14d: #3985, approximately 1.2 days old. Eligible App-authored PR #3989, head docs/drafted-solutions, approximately 0.7 days / 16 hours old, explicitly Closes #3985. Zero uncovered proposals, so neither stall rule fires. The nonempty harvest/draft/publish run succeeds; the expected docs extension exists in that PR. Human merge remains pending. Metrics #3674 is not proposal-authorship evidence.

Degraded review consumer: #3989 has zero reviews/comments and still requests fro-bot and marcusrbrown. Both green first and second Fro Bot jobs log unauthorized_author: association CONTRIBUTOR is unauthorized, before any model review. The workflow’s narrow drafted-branch exception admits the job, but the pinned v0.119.2 action independently rejects the association. Publication is verified; review acceptance is absent. Do not call this an unauthored/stalled proposal queue or a successful automated review.

Needs Human Attention

1. Drafted-doc review admission: evidence is #3989 and its linked rejected jobs. Exact paths: control-plane .github/workflows/fro-bot.yaml:485–518,618, .github/workflows/draft-solutions.yaml, scripts/drafted-solutions-publish.ts; upstream fro-bot/agent/src/features/triggers/router.ts, skip-conditions-pr.ts, author-utils.ts and paired tests. Root cause: workflow admission and action authorization disagree for the App-triggered review. Smallest safe next step: inspect sender/association resolution and reconcile an exact same-repository App + drafted-head + verified review-request trust path, or arrange an authorized human review now. Do not authorize all CONTRIBUTORS/bots, bypass fork/privacy gates, or repeatedly request the same rejected review. Verify negative spoof/fork/untrusted-sender tests and an actual review on the current head, not green checks. Prior learning: docs/solutions/best-practices/agent-drafts-trusted-publish-split-2026-10-10.md, especially its exact-predicate rule; token wiring alone cannot establish event trust.

2. Rollout reconciliation: dedicated tracker owns #3512. Paths: agent packages/gateway/src/operator-contract/version.ts, dashboard src/gateway/operator-contract/version.ts, infra apps/gateway/upstream.json, apps/dashboard/docker-compose.yaml, docs/runbooks/gateway-operator-auth-lifecycle.md. Reconcile six body Open / Project Done values and stale body/ledger versions from actual deploy evidence. Do not adopt a 1.10.0 gateway against the 1.9.0 dashboard or weaken exact matching. Verify authenticated acceptance and retain paired volume-aware rollback.

3. Guard and least privilege — remediation handoff: reuse #3979; inspect effective redacted CI OpenCode initialization/configuration and .github/workflows/fro-bot.yaml. Unavailable guard has no established root cause; do not retry unavailable start/complete calls or widen credentials. Separately verify pinned CodeQL requirements before removing surplus packages: read/actions: read from .github/workflows/codeql-analysis.yaml:27–31; preserve scan coverage and validate required commands/YAML/action refs plus a CodeQL run. Prior learning: docs/solutions/workflow-issues/required-github-token-for-agent-steps-2026-06-22.md distinguishes wiring from scope.

4. Fleet coverage and CI: security HTTP statuses and empty check/workflow sources are detection limits. Confirm intended scanner/CI coverage on each linked repository; inspect failed-step inputs before retry and reuse existing security PRs/dashboards. Routine upgrades stay Renovate-owned. Reuse #3793 for a Renovate-preset validation gate through shared setup; preserve all existing guards.

5. Durable handoff: additive changes are limited to knowledge/wiki/topics/github-actions-ci.md, knowledge/index.md, append-only knowledge/log.md. Prior overlay/rename changes and staged metadata remain preserved; this pass made no metadata edit. The wiki diff remains dirty for trusted caller ingestion; persistence is unverified. Existing unrelated broken links remain #3903. The sole response comment also retires the prior daily report; no per-finding issue/PR comments, tracker writes, branch/commit/push or PR mutations.

Run 38355617420 · oversight working-dir · caller owns trusted knowledge ingestion.