refactor(ci): one GitHub API client, one sticky PR-comment action and one schema-split rule set (#15411)

[mrveiss]

Thinking Path

Carried from three stale local branches (issue-15411-github-api-module, issue-12984-pr-comment-upsert, issue-18158-tooling-clones) under the owner rule "any sitting branch must be linked to a finished PR" (branch cleanup 2026-10-12). All are CI/pipeline tooling. Built fresh from current main, with only each branch's unique commits cherry-picked; no stale tree was revived.

What Changed

• tech-debt(ci): GitHubApi is shared by three modules but lives inside a size-frozen one, so the shared client can no longer gain a method #15411, one GitHub API client:
  • GitHubApi is extracted from pipeline-scripts/ci_dispatch_watchdog.py into pipeline-scripts/github_api.py. The three importers (ci_dispatch_retry.py, ci_red_cause.py, release_sync_main.py) use it.
  • rerun_run and rate_limit_remaining move onto GitHubApi too.
  • The watchdog drops 1370 → 1244 lines (ceiling lowered in both size records).
  • The egress baseline row moves from the watchdog to github_api.py, with the total unchanged.
• refactor(ci): converge two diverged PR-comment upserts; ssot-coverage appends duplicates past 30 comments #12984, one sticky PR comment: a composite action, .github/actions/upsert-pr-comment/, that is marker-based and paginated. ssot-coverage.yml and ai-security-review.yml use it, and the inline upsert is removed. ci(perf): collapse 15 duplicate paths-filter jobs into one reusable workflow #15360's changes to ai-security-review.yml are kept; only the upsert step differs. The five fire-and-forget comment callers are untouched. env_registry_testing.py and ENV_VARS.md gain one row.
• tech-debt(tooling): two script pairs clone their rules table and CLI preamble — found by the widened duplication scan #18158 / tech-debt(tooling): tools/schema-split/resolve_schema_conflicts.py still writes output with print() #18170, tooling clones:
  • tools/schema-split/schema_rules.py (one rule set for the schema-split tools);
  • scripts/drift_cli.py (one drift CLI);
  • tools/schema-split/resolve_schema_conflicts.py uses the shared logging, with no print(.
• For the reviewer, an egress policy call: repo_tests/egress_default_guard_17576_test.py EXEMPT gains .github/actions/upsert-pr-comment/upsert_pr_comment.py.
  • It runs on a bare runner with the standard library only, so it cannot import the guard, and the detector never counts urlopen as guarded.
  • It talks only to the runner's own GitHub API root with the job token, never a user-, LLM- or document-supplied URL.
  • The alternative, raising the drop-only baseline, is forbidden. If you'd rather scope the whole .github/ tree out of the AutoBot-runtime egress census, say so.

Verification

• Cherry-picks: tech-debt(ci): GitHubApi is shared by three modules but lives inside a size-frozen one, so the shared client can no longer gain a method #15411 d45f6c11ef → caba5d6; refactor(ci): converge two diverged PR-comment upserts; ssot-coverage appends duplicates past 30 comments #12984 71be52b → db9f1e6 and 6366385 → 205f9a5; tech-debt(tooling): two script pairs clone their rules table and CLI preamble — found by the widened duplication scan #18158 5cab0c6f66 → 78fbe67; tech-debt(tooling): tools/schema-split/resolve_schema_conflicts.py still writes output with print() #18170 2af14d1be2 → eca5071.
  • Conflicts only in repo_tests/_reach_policy.py and reach_scope_claim_17844_test.py (hub files), where main's entries are kept plus each commit's own.
• Reversal check: no main commit since the 2026-10-10 merge base is undone.
  • The tool and script files have no main changes.
  • In ai-security-review.yml, ci(perf): collapse 15 duplicate paths-filter jobs into one reusable workflow #15360 (0a3d672) is kept.
  • Hub files: only additions, plus the lowered ceiling.
• Tests, one run on the final head: the 8 changed/added test files + egress_default_guard_17576 + reach_declarations + reach_scope_claim_17844 + the size ratchet: 738 passed.
• Not run: actionlint, which isn't installed locally; CI runs the workflow lint.

Risks

• The sticky-comment action: a behaviour change for two workflows' PR comments (upsert instead of inline logic). Same marker semantics.

Model Used

Anthropic Claude Opus 5.5 (claude-opus-5-5), coordinating. Anthropic Claude Sonnet 5.5 (claude-sonnet-5-5) subagent for the cherry-picks and verification.

Issue Link

Closes #18170
Refs #15411
Refs #12984
Refs #18158

Still open:

• tech-debt(ci): GitHubApi is shared by three modules but lives inside a size-frozen one, so the shared client can no longer gain a method #15411: the optional CAUSE_* → StrEnum item;
• refactor(ci): converge two diverged PR-comment upserts; ssot-coverage appends duplicates past 30 comments #12984: until CI's workflow lint passes and the egress exemption above is ruled on;
• tech-debt(tooling): two script pairs clone their rules table and CLI preamble — found by the widened duplication scan #18158 AC3: the duplication-guard pin drop (.github/workflows/duplication-guard.yml ~:94, 11707) needs a jscpd measurement, which can't run locally while tooling(slm-frontend): the shared SLM node_modules was emptied and now holds a link to itself — every worktree's SLM vue-tsc and vitest are dead #18406 is open.

Single-issue rationale: one CI-tooling scope carried from three stale branches. Only #18170 is fully delivered; the other three are partial and stay open, as listed above.

Changelog fragment

• N/A (CI/pipeline tooling)

[coderabbitai]

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 59 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Repository: mrveiss/AutoBot-AI/.coderabbit.yaml
• Review profile: ASSERTIVE
• Plan: Advanced
• Run ID: 5a9d4e52-d741-408a-bf4b-bc263954376f

---

📥 Commits

Reviewing files that changed from the base of the PR and between df211d3 and e968902.

---

⛔ Files ignored due to path filters (2)

• repo_tests/python_file_size_ratchet_baseline.py is excluded by !repo_tests/python_file_size_ratchet_baseline.py
• scripts/python_file_size_known_large.py is excluded by !scripts/python_file_size_known_large.py

---

📒 Files selected for processing (30)

• .github/actions/upsert-pr-comment/action.yml
• .github/actions/upsert-pr-comment/upsert_pr_comment.py
• .github/workflows/ai-security-review.yml
• .github/workflows/ssot-coverage.yml
• autobot_shared/env_registry_testing.py
• docs/developer/ENV_VARS.md
• pipeline-scripts/ci_dispatch_retry.py
• pipeline-scripts/ci_dispatch_watchdog.py
• pipeline-scripts/ci_red_cause.py
• pipeline-scripts/github_api.py
• pipeline-scripts/release_sync_main.py
• repo_tests/_reach_policy.py
• repo_tests/ci_dispatch_retry_test.py
• repo_tests/egress_default_guard_17576_test.py
• repo_tests/egress_unguarded_baseline_17576.json
• repo_tests/github_api_behaviour_15411_test.py
• repo_tests/github_api_single_module_15411_test.py
• repo_tests/pr_comment_upsert_12984_test.py
• repo_tests/reach_scope_claim_17844_test.py
• scripts/backfill_relationships.py
• scripts/drift_cli.py
• scripts/drift_cli_test.py
• scripts/scope_box_artefact_drift.py
• scripts/umbrella_label_drift.py
• tools/schema-split/README.md
• tools/schema-split/resolve_schema_conflicts.py
• tools/schema-split/schema_rules.py
• tools/schema-split/schema_rules_test.py
• tools/schema-split/split_schemas.py
• tools/schema-split/update_imports.py

---

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Notice: 29 open PRs — past the runaway threshold (25)

This is not a request to defer this PR. The actual limit is the open-PR cap (AUTOBOT_OPEN_PR_CAP, default 40), enforced at pre-push against NEW branches only -- pushing more commits to this PR's own branch is never blocked by it.

This notice only means the count is high enough to be worth a glance for a runaway — something opening PRs in a loop, or a merge pipeline that has stalled so nothing is draining.

Currently open:

• refactor(forks): resolve the agents-kb symbol-fork batch (#18457) #18465 refactor(forks): resolve the agents-kb symbol-fork batch (tech-debt(forks): resolve the 'agents-kb' batch of symbol-fork baseline clusters — merge real forks, rename homonyms apart #18457)
• refactor(forks): resolve the api batch of symbol-fork clusters, 13 of 13 names (#18451) #18463 refactor(forks): resolve the api batch of symbol-fork clusters, 13 of 13 names (tech-debt(forks): resolve the 'api' batch of symbol-fork baseline clusters — merge real forks, rename homonyms apart #18451)
• feat(kb): one provenance read accessor over the five ingestion dialects, the dialect readers migrated, and a guard (#17618) #18462 feat(kb): one provenance read accessor over the five ingestion dialects, the dialect readers migrated, and a guard (bug(llc): the KB context's sources answers tenancy, not where the document came from #17618)
• refactor(forks): resolve the 'services' symbol-fork batch (#18452) #18460 refactor(forks): resolve the 'services' symbol-fork batch (tech-debt(forks): resolve the 'services' batch of symbol-fork baseline clusters — merge real forks, rename homonyms apart #18452)
• refactor(forks): resolve the utils-models symbol-fork batch, 18 clusters (#18450) #18455 refactor(forks): resolve the utils-models symbol-fork batch, 18 clusters (tech-debt(forks): resolve the 'utils-models' batch of symbol-fork baseline clusters — merge real forks, rename homonyms apart #18450)
• feat(automation): three AI lines per tool action behind one approval enforcement point (#18430, #17463) #18453 feat(automation): three AI lines per tool action behind one approval enforcement point (feat(automation): three AI lines per tool action — autonomous, AI drafts and a person sends, never touches — and a four-move automation setup #18430, tech-debt(approvals): three approval policies, one enforced — needs_approval() and the 'minimal' guard profile have no production caller #17463)
• fix(chat): cancel the graph turn on disconnect, bound it in time, add a turn-cancel route (#17273) #18447 fix(chat): cancel the graph turn on disconnect, bound it in time, add a turn-cancel route (bug(chat): a disconnected client leaves the LangGraph turn running, and the LLM call has no total timeout #17273)
• feat(company-os): roles, workflows and processes data model with company-scoped API (#18390) #18444 feat(company-os): roles, workflows and processes data model with company-scoped API (feat(company-os): roles own workflows of reusable processes — transitions, triggers, skills, meta skills, routine-run processes #18390)
• feat(chat): Refined UI step 5a — Chat tabs as a tablist, one-at-a-time side panels, presence sync; design docs for #18238 (#18243) #18441 feat(chat): Refined UI step 5a — Chat tabs as a tablist, one-at-a-time side panels, presence sync; design docs for docs(design): refresh the Refined UI handoff and mockups for the SLM split, BI move and restricted Usage #18238 (Refined UI step 5a · Chat #18243)
• chore(deps): bump the all-dependencies group across 1 directory with 5 updates #18440 chore(deps): bump the all-dependencies group across 1 directory with 5 updates
• chore(deps): bump the all-dependencies group across 1 directory with 18 updates #18439 chore(deps): bump the all-dependencies group across 1 directory with 18 updates
• chore(deps): bump the all-dependencies group across 1 directory with 23 updates #18438 chore(deps): bump the all-dependencies group across 1 directory with 23 updates
• chore(deps): bump the all-dependencies group across 1 directory with 10 updates #18437 chore(deps): bump the all-dependencies group across 1 directory with 10 updates
• chore(deps): bump the all-dependencies group across 1 directory with 3 updates #18436 chore(deps): bump the all-dependencies group across 1 directory with 3 updates
• chore(deps): bump the all-dependencies group across 1 directory with 6 updates #18435 chore(deps): bump the all-dependencies group across 1 directory with 6 updates
• chore(deps): bump the all-dependencies group across 1 directory with 3 updates #18434 chore(deps): bump the all-dependencies group across 1 directory with 3 updates
• chore(deps): bump the all-dependencies group across 1 directory with 3 updates #18433 chore(deps): bump the all-dependencies group across 1 directory with 3 updates
• fix(egress): guard the integrations under SSOT policies, tenant-configured URLs public-only, websockets refuse redirects (#17576) #18429 fix(egress): guard the integrations under SSOT policies, tenant-configured URLs public-only, websockets refuse redirects (security(egress): rule 8's only enforcement is a five-file literal, and four connectors beside it call tracked_request unguarded #17576)
• feat(backup): ciphertext-only secrets-store backup through the SLM backup service, and one fail-closed verify_backup (#18281) #18425 feat(backup): ciphertext-only secrets-store backup through the SLM backup service, and one fail-closed verify_backup (feat(backup): the secrets store has no backup mechanism, so destructive secrets proposals cannot verify a backup_reference #18281)
• feat(slm): node-major post-sync plan with a durable per-action record, sudo -n restarts and a node sync endpoint (#17455) #18423 feat(slm): node-major post-sync plan with a durable per-action record, sudo -n restarts and a node sync endpoint (feat(sync): post-sync actions need a durable record and a node-level plan in the orchestrator #17455)
• refactor(rag): de-fork graph retrieval — one neighbour contract, one graph blend, shared config (#18139, #18141) #18421 refactor(rag): de-fork graph retrieval — one neighbour contract, one graph blend, shared config (bug(rag): PPR is wired with the mesh adapter whose get_neighbors signature differs — mesh retrieval raises TypeError on first use #18139, tech-debt(rag): graph-blended retrieval paths fork the score blend, config, reranker weights and stale filter #18141)
• refactor(captcha): one CaptchaType in autobot_shared (#18131) #18418 refactor(captcha): one CaptchaType in autobot_shared (tech-debt(captcha): two captcha_solver modules whose CaptchaType enums disagree — a kind one can classify is UNKNOWN to the other #18131)
• fix(memory): host context calls get_related_entities, the canonical graph traversal (#18140) #18416 fix(memory): host context calls get_related_entities, the canonical graph traversal (bug(memory): security memory integration calls traverse_relations, which the memory graph does not have #18140)
• feat(guard): a kit component cannot be forked in either app, and kit stories render in both Storybooks (#14776) #18409 feat(guard): a kit component cannot be forked in either app, and kit stories render in both Storybooks (Retire the local Base* component forks in autobot-frontend into the shared kit #14776)
• feat(agents): read-only agent-ledger APIs for Company OS and the SLM (#18391) #18408 feat(agents): read-only agent-ledger APIs for Company OS and the SLM (feat(agents): a GUI for the agent ledger — who is working on what, per company and install-wide in the SLM #18391)
• fix(kb): add_document chunks, partial-success ingest, stable upload rejection codes (#17830) #18405 fix(kb): add_document chunks, partial-success ingest, stable upload rejection codes (bug(kb): add_document never chunks — a whole transcript or PDF becomes one embedding, against its own docstring #17830)
• feat(approvals): one approval record, shared RiskTier, pending inbox query (#13421 S1) #18404 feat(approvals): one approval record, shared RiskTier, pending inbox query (ui(approvals): five approval surfaces with no shared record or component — build one approval inbox #13421 S1)
• feat(llm): needs_reauth registry skip + typed error, header headroom harvest, pre-dispatch headroom routing (#15022, #15026, #15028) #18403 feat(llm): needs_reauth registry skip + typed error, header headroom harvest, pre-dispatch headroom routing (llm: degradation store cannot distinguish a dead credential from a transient failure — dead tokens retry every TTL forever #15022, llm: harvest provider rate-limit response headers into a cross-worker quota headroom store (the missing quota monitor) #15026, llm: route on observed headroom before dispatch instead of on a 429 after it; reconcile guessed RPM buckets with observed limits #15028)
• fix(slm-auth): SSO callbacks redirect with a one-time code, not the session JWT (#18394) #18399 fix(slm-auth): SSO callbacks redirect with a one-time code, not the session JWT (security(sso): SSO callbacks redirect with the session JWT in the query string — now live after #18322 #18394)

If the queue is draining normally, ignore this. Otherwise:

1. Check whether CI is dispatching at all — see the ci-dispatch-watchdog status on these PRs
2. Merge the ones whose CI has finished and review has passed: gh pr merge <number> --squash --delete-branch
3. Look for a loop opening near-identical PRs

Warn-only runaway detector — .github/workflows/pr-queue-gate.yml. It never blocks a merge.

[mrveiss]

Security review (Opus) at e968902: approve.

• Reversal check: clean. The merge-base is current main (0 behind), and for all 32 files the two-dot and three-dot line counts match, so ci(perf): collapse 15 duplicate paths-filter jobs into one reusable workflow #15360s ai-security-review.yml` changes are kept.
• No script injection in the new composite action: every input reaches the shell only via env: as "$VAR", and the body travels as a file path (action.yml:46-55).
• Triggers: both workflows run on pull_request, not pull_request_target, so a fork gets a read-only token, the comment 403s, and continue-on-error absorbs it.
• Upsert script: the body is sent only as a JSON payload; the token is in a header and never logged; the API root comes from github.api_url.
• github_api.py: the token is in a header, and transport errors carry no headers.

Policy decision (merger): option (a), the per-file EXEMPT for upsert_pr_comment.py, as written. It is stdlib-only on a bare runner, cannot import the guard, and talks only to the runners API root with the job token. Taking all of .github/out of the census would silently exclude any future CI script that fetches a PR- or LLM-supplied URL. The entry carries its reason in a comment (#12984). A reason-enforcingdictforEXEMPT` would be stronger, but is optional.

Pre-existing items, not this PR`s, filed as #18468:

• github_api.py:82 sends the token to any absolute URL;
• ssot-coverage.yml still interpolates step outputs into run:, the security(ci): five workflow expression-injection sites interpolate untrusted context directly into run: bodies #13516 class;
• the upsert script accepts an http:// root;
• a guard for the class.

Queued for the next merge train.

[github-actions]

✅ SSOT Configuration Compliance: Passing

🎉 No new hardcoded values of either class — ssot and other both block.

Known backlog in pipeline-scripts/hardcoded_values_baseline.txt is suppressed and tracked in #14371.