chore: sync workflow templates

[stranske]

Sync Summary

Files Updated

• agents-guard.yml: Agents guard - enforces agents workflow protections (Health 45)
• check_agents_md_freshness.py: Warns when the generated Orchestrator AGENTS.md playbook section cites stale repo paths or commands
• runner_lib/ (1 files): Shared runner prompt assembly, output parsing, and dispatch debounce helpers
• langchain_client.py: LangChain client builder - multi-provider client with slot-based fallback and configuration
• llm_registry.py: LLM model registry helper - shared slot/model selection and blocked-model enforcement

Files Skipped

• pr-00-gate.yml: File exists and sync_mode is create_only
• ci.yml: File exists and sync_mode is create_only
• renovate.json: File exists and sync_mode is create_only
• cross-repo-smoke.yml: File exists and sync_mode is create_only
• .github/scripts/package.json: Repo installs .github/scripts dependencies from package-lock.json and forbids tracked node_modules
• .github/scripts/node_modules/minimatch: Repo installs .github/scripts dependencies from package-lock.json and forbids tracked node_modules
• .github/scripts/node_modules/brace-expansion: Repo installs .github/scripts dependencies from package-lock.json and forbids tracked node_modules
• .github/scripts/node_modules/balanced-match: Repo installs .github/scripts dependencies from package-lock.json and forbids tracked node_modules
• llm_slots.json: None

---

Review Checklist

• CI passes with updated workflows
• No repo-specific customizations were overwritten

Source: stranske/Workflows
Source SHA: 821168725d0568cade5bb9a18fc58cab97657975
Template hash: 591316374281
Sync branch: sync/workflows-591316374281
Consumer repo: stranske/trip-planner
Manifest: .github/sync-manifest.yml

Summary by CodeRabbit

• Bug Fixes

  • Improved error extraction and reporting from runner output, ensuring clearer error messages when operations fail
  • Enhanced model selection fallback logic to better handle blocked overrides and provide more reliable provider selection

• Refactor

  • Strengthened path validation and resolution across internal components to improve security and reliability

[netlify]

✅ Deploy Preview for stranske-trip-planner canceled.

Name	Link
🔨 Latest commit	5900bff
🔍 Latest deploy log	https://app.netlify.com/projects/stranske-trip-planner/deploys/6a38ca165f5ac700083320cb

[coderabbitai]

Caution

Review failed

Pull request was closed or merged during review

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: c339016e-34cf-4e18-adc2-e59284e75815

📥 Commits

Reviewing files that changed from the base of the PR and between 615d0c6 and 5900bff.

📒 Files selected for processing (5)

• .github/workflows/agents-guard.yml
• scripts/check_agents_md_freshness.py
• scripts/runner_lib/core.py
• tools/langchain_client.py
• tools/llm_registry.py

🔗 Linked repositories identified

CodeRabbit considers these linked repositories for cross-repo context during reviews:

• stranske/Workflows (auto-detected)
• stranske/Template (auto-detected)

📜 Recent review details

⏰ Context from checks skipped due to timeout. (2)

• GitHub Check: guard
• GitHub Check: cross-repo-full-product

🧰 Additional context used

📓 Path-based instructions (7)

{pyproject.toml,.github/workflows/**/*.{yml,yaml}}

📄 CodeRabbit inference engine (.github/copilot-instructions.md)

Check both pyproject.toml ([tool.coverage.report] fail_under) AND workflow files for coverage-min settings; ensure these match or the lower value will be the effective threshold

Files:

• .github/workflows/agents-guard.yml

.github/workflows/**/*.{yml,yaml}

📄 CodeRabbit inference engine (.github/copilot-instructions.md)

For GitHub Actions workflow startup_failure errors with zero jobs, check for invalid YAML syntax, top-level permissions: block on workflow_call reusable workflows (which conflicts with caller permissions), invalid permission scopes, or circular workflow references

Files:

• .github/workflows/agents-guard.yml

.github/workflows/*.yml

📄 CodeRabbit inference engine (CLAUDE.md)

First-party consumers should reference reusable workflows with @main unless intentionally pinning to an exact commit SHA for a controlled reason

Files:

• .github/workflows/agents-guard.yml

.github/workflows/agents-*.yml

📄 CodeRabbit inference engine (CLAUDE.md)

agents-*.yml, autofix.yml, .github/codex/ prompts, and synced scripts/docs should be fixed in stranske/Workflows, not in the consumer repo

Files:

• .github/workflows/agents-guard.yml

**/.github/workflows/*.yml

📄 CodeRabbit inference engine (AGENTS.md)

First-party consumers should reference reusable workflows with @main unless intentionally pinning to an exact commit SHA for a documented controlled reason

Files:

• .github/workflows/agents-guard.yml

{.github/workflows/agents-*.yml,.github/workflows/autofix.yml,.github/codex/**,docs/**}

📄 CodeRabbit inference engine (AGENTS.md)

Synced workflows (agents-*.yml, autofix.yml), prompts in .github/codex/, and synced scripts/docs should be fixed in stranske/Workflows, not locally; do not edit them locally

Files:

• .github/workflows/agents-guard.yml

**/*.py

📄 CodeRabbit inference engine (.github/copilot-instructions.md)

In the Manager-Database repository (stranske/Manager-Database), use Prefect 2.x and import schedules from prefect.client.schemas.schedules

Files:

• tools/langchain_client.py
• scripts/runner_lib/core.py
• tools/llm_registry.py
• scripts/check_agents_md_freshness.py

🪛 ast-grep (0.43.0)

scripts/runner_lib/core.py

[info] 557-557: use jsonify instead of json.dumps for JSON output
Context: json.dumps(event, sort_keys=True)
Note: Security best practice.

(use-jsonify)

🔀 Multi-repo context stranske/Workflows, stranske/Template

Perfect. I have completed my cross-repository exploration and have sufficient information to conclude.

Linked repositories findings

stranske/Workflows (Source Template Repository)

Key Consumers of Changed APIs:

• [::stranske/Workflows::] scripts/langchain/_llm_client.py:35-41 — Wraps build_chat_client() to provide import-guarded construction for agent scripts (progress_reviewer, followup_issue_generator, pr_verifier, etc.). This wrapper transparently inherits the new "skip + fallback" override behavior.
• [::stranske/Workflows::] scripts/langchain/followup_issue_generator.py:24-40 — Uses build_chat_client with reasoning-model selection
• [::stranske/Workflows::] scripts/langchain/progress_reviewer.py:18-22 — Uses build_chat_client for LLM-powered review

Critical Behavioral Change: Model Override Fallback Strategy

• [::stranske/Workflows::] tools/langchain_client.py — Changed from pre-loop early refusal to in-loop per-slot fallback. Old behavior: if a user-requested model override is blocked, return None immediately. New behavior: skip that slot and continue to next slot; if override was requested but the original slot's model is safe, keep the original instead of dropping it.
• [::stranske/Workflows::] tools/llm_registry.py:85-104 — apply_slot_env_overrides() now distinguishes "override explicitly requested" from "override was blocked." If override requested and blocked, but original slot is safe: keeps original. If override blocked and original also blocked: skips the slot entirely.

Comprehensive Test Coverage for Behavior Changes:

• [::stranske/Workflows::] tests/tools/test_langchain_client.py:124-145 — Test test_build_chat_client_blocked_model_override_does_not_shift_provider validates the new behavior: blocked global overrides are NOT retried against other provider slots
• [::stranske/Workflows::] tests/tools/test_langchain_client.py:200-230 — Test test_apply_slot_env_overrides_keeps_original_when_override_blocked_but_original_safe validates slot-level: keeps original when override blocked but original safe

Enhanced Path Safety in Runner Library:

• [::stranske/Workflows::] scripts/runner_lib/core.py — Added safe path-resolution helpers to prevent directory traversal during orchestrator-skill materialization. Validates destination_root and checkout_path resolve within workspace boundaries, rejecting .. patterns.
• [::stranske/Workflows::] tests/scripts/test_runner_lib.py:170-185 — Test test_materialize_orchestrator_skill_rejects_unsafe_pack_checkout with parametrized paths .reference/. and .reference/.. confirms unsafe paths are rejected

Safer Command Ref Parsing:

• [::stranske/Workflows::] scripts/check_agents_md_freshness.py — Enhanced with shlex.split() for shell-like quoting support (vs. naive whitespace split). New _clean_ref() strips only matched paired quotes. New _resolve_repo_path() validates paths stay within repo_root.

Sync Manifest Confirms Delivery:

• [::stranske/Workflows::] .github/sync-manifest.yml:285-305 — Confirms all changed modules are marked delivery: copy (copy-synced to consumer repos):
  • tools/langchain_client.py — "LangChain client builder - multi-provider client with slot-based fallback and configuration"
  • tools/llm_registry.py — "LLM model registry helper - shared slot/model selection and blocked-model enforcement"
  • scripts/runner_lib/ — "Shared runner prompt assembly, output parsing, and dispatch debounce helpers"
  • scripts/check_agents_md_freshness.py — "Warns when the generated Orchestrator AGENTS.md playbook section cites stale repo paths or commands"

stranske/Template (Secondary Template Reference)

• [::stranske/Template::] Same modules present and used identically by consumer scripts (scripts/langchain/followup_issue_generator.py, scripts/langchain/_llm_client.py)

Impact Assessment

Backwards Compatibility: The changes are backwards-compatible. Consumers that relied on early refusal of blocked overrides will now gracefully fall back to safe alternatives or original slots, which is a strict improvement.

Safety Hardening: All three enhancement areas (override fallback logic, path resolution validation, ref parsing) are validated with comprehensive test coverage in the source repository.

🔇 Additional comments (14)

tools/llm_registry.py (1)

82-88: LGTM!

Also applies to: 97-104, 268-272

tools/langchain_client.py (1)

283-301: LGTM!

scripts/check_agents_md_freshness.py (5)

9-9: LGTM!

---

50-55: LGTM!

---

65-97: LGTM!

---

100-111: LGTM!

---

167-172: LGTM!

scripts/runner_lib/core.py (6)

87-112: LGTM!

---

346-346: LGTM!

---

397-400: LGTM!

---

461-466: LGTM!

---

551-558: LGTM!

---

576-576: LGTM!

.github/workflows/agents-guard.yml (1)

114-114: LGTM!

Also applies to: 183-183

---

📝 Walkthrough

Walkthrough

The PR hardens path resolution in runner_lib/core.py and check_agents_md_freshness.py against path traversal, improves shell-ref parsing with shlex, fixes LLM model-override blocking in langchain_client.py to fall back per-slot instead of aborting, adds defensive validation to llm_registry.py's payload parsing, and bumps a pinned CI action SHA.

Changes

Safe path resolution and shell-ref hardening

Layer / File(s)	Summary
check_agents_md_freshness.py: shlex parsing and root-validated path checks scripts/check_agents_md_freshness.py	Adds shlex import; tightens _clean_ref() to strip only matching surrounding quotes; introduces _command_parts() and _resolve_repo_path() for shell-safe tokenization and root-confined path resolution; updates _command_exists() and _check_command_ref() to use these helpers; resolves --agents-md against repo_root in main().
runner_lib/core.py: safe child-path helpers and checkout materialization scripts/runner_lib/core.py	Introduces _resolve_child_path and _resolve_reference_checkout_path to enforce paths remain within directory roots; updates _materialize_single_checkout_plan, materialize_orchestrator_skill, and assemble_prompt to route path computations through these validators.
runner_lib/core.py: JSONL error extraction and final-message selection scripts/runner_lib/core.py	_parse_jsonl_output now handles dict-shaped and string-shaped event["error"] payloads, appending the most specific error text; parse_runner_output prefers the first parsed error as final_message when errors are present.

LLM registry validation and model-override fallback

Layer / File(s)	Summary
Registry payload validation and blocked-override slot retention tools/llm_registry.py	load_model_registry() verifies payload["models"] is a list and each entry's quality field is a dict before parsing, logging warnings on malformed data. apply_slot_env_overrides() appends the original slot when an override was requested but blocked and the original slot is not blocked.
Per-slot override blocking and fallback in build_chat_client tools/langchain_client.py	Moves model-override handling into the slot iteration loop; adds per-slot provider availability checks; uses a used_override flag to apply the override once, logs a warning on a blocked override, and falls back to the slot's default model rather than returning None.

CI workflow SHA bump

Layer / File(s)	Summary
agents-guard.yml: bump setup-api-client SHA .github/workflows/agents-guard.yml	Replaces the old commit SHA with c2537cc959f2ce05926c4639d25b90678abc97bc for the stranske/Workflows/.github/actions/setup-api-client reference in both the pull_request_target and pull_request fallback steps.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• stranske/trip-planner#1430: Also updates the pinned stranske/Workflows/.github/actions/setup-api-client SHA in the same two fallback steps of agents-guard.yml.
• stranske/trip-planner#1431: Also modifies scripts/runner_lib/core.py for orchestrator-skill materialization and orchestrator-skill context incorporation into prompt assembly, directly overlapping with the path-handling changes in this PR.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 15.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'chore: sync workflow templates' accurately reflects the main objective of the pull request, which is to synchronize workflow templates and related configuration files from the source repository.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch sync/workflows-591316374281

---

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