smykla-skalski/gh-review-comments
codex/gh-review-comments/SKILL.md
List, reply to, resolve, unresolve, and create GitHub PR review comment threads using gh CLI scripts.
$ install --global
skillsauthnpx skillsauth add smykla-skalski/sai gh-review-commentsInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 26, 2026, 8:05 AM43.8s8 files scanned
SKILL.md
- name:
- gh-review-comments
- description:
- List, reply to, resolve, unresolve, and create GitHub PR review comment threads using gh CLI scripts.
GH Review Comments
Use this skill when the user wants to manage pull request review threads on GitHub (list, reply, resolve, unresolve, or create a review with inline comments).
Inputs to infer from natural-language requests
Extract these values from the user request or current repository context:
owner/repo(required)- PR number (required)
- optional reviewer filter:
--author <login> - optional target thread:
--thread-id <PRRT_...> - optional action details:
--reply <message>--resolve--unresolve--create-review--unresolved-only
If action details are missing, default to listing threads first and ask a focused follow-up question.
Auth and sandbox preflight (Codex-specific)
Before any script call:
- Run
gh auth status. - If the command fails due sandbox/network restrictions, rerun the same command with
sandbox_permissions=require_escalatedand a short justification. - If auth is not valid, ask the user to authenticate with
gh auth login, then re-rungh auth status. - If a later
gh apicall fails due sandbox/network restrictions, rerun that command with escalation.
Script directory
Use this concrete skill path:
SKILL_DIR="/Users/[email protected]/Projects/github.com/smykla-skalski/sai/codex/gh-review-comments"
Run scripts from that directory so behavior is consistent:
"$SKILL_DIR/scripts/<script>.sh" [args...]
Scripts
scripts/list-threads.sh
"$SKILL_DIR/scripts/list-threads.sh" <owner> <repo> <pr_number> [--author <login>] [--unresolved-only]
Output: one JSON line per thread with thread_id, comment_id, author, body, path, line, is_resolved, is_outdated, reply_count.
scripts/reply-thread.sh
"$SKILL_DIR/scripts/reply-thread.sh" <owner> <repo> <pr_number> <comment_id> <body>
comment_idmust be the integer top-level comment ID fromlist-threads.sh.
scripts/resolve-thread.sh
"$SKILL_DIR/scripts/resolve-thread.sh" <thread_id>
thread_idmust be the GraphQL node ID (PRRT_...) fromlist-threads.sh.
scripts/unresolve-thread.sh
"$SKILL_DIR/scripts/unresolve-thread.sh" <thread_id>
scripts/create-review.sh
"$SKILL_DIR/scripts/create-review.sh" <owner> <repo> <pr_number> <event> <body> [<comments_json>|-]
event:COMMENT,APPROVE, orREQUEST_CHANGEScomments_json: array of objects withpath,line,body, optionalside(defaults toRIGHT)- pass
-to read comments JSON from stdin
Workflow
1. Interpret intent
Map user request to one of:
- List threads
- Reply to threads
- Resolve threads
- Reply then resolve
- Unresolve threads
- Create review with inline comments
For reply/resolve operations, enable unresolved filtering unless the user explicitly asks otherwise.
2. List threads
- Run
list-threads.shwith required args and any filters. - If
--thread-idis provided, filter to that thread. - Present thread summary (id, author, file/line, resolution status, short body preview, reply count).
If no action is requested, stop here.
3. Execute thread actions
For each matched thread:
- Reply: use
comment_idwithreply-thread.sh - Resolve: use
thread_idwithresolve-thread.sh - Unresolve: use
thread_idwithunresolve-thread.sh - Reply + Resolve: run reply first, then resolve
Continue processing remaining threads if one fails; report partial failures clearly.
4. Create review (alternative path)
When the user asks to create a review:
- Collect
event, review body, and inline comments. - Validate each inline comment target against files/lines in the PR diff.
- Run
create-review.shexactly once. - If it exits 0, treat it as success for the full review payload.
Do not automatically retry create-review.sh after a successful run because that creates duplicate reviews.
5. Verify and summarize
- For reply/resolve/unresolve: re-run
list-threads.shand compare before/after states. - For create-review: use
create-review.shsuccess output as source of truth.
Report:
- total matched threads
- actions executed
- verification results (successes/failures)
- relevant GitHub URLs
Error handling
- If
--thread-idmatches nothing, report no matching thread. - If one thread operation fails, continue with remaining matches and summarize failures.
- If
create-review.shfails, report error and likely cause (for example invalidpath/linetargets).
Read references/gh-api-guide.md before taking action to avoid ID mismatches between REST integer IDs and GraphQL node IDs.
Example user requests (Codex style)
- "List all review threads on
smykla-skalski/saiPR4." - "Show only unresolved threads by
Automaatonsmykla-skalski/saiPR4." - "Reply
Fixed in latest pushto unresolved threads byAutomaaton PR4." - "Resolve thread
PRRT_kwDOCnTGG85tgSD3onsmykla-skalski/saiPR4." - "Reply
Done, thanks!and resolve threadPRRT_kwDOCnTGG85tgSD3on PR4." - "Create a
REQUEST_CHANGESreview onsmykla-skalski/saiPR4with inline comments."
Related Skills
smykla-skalski/council
development
VerifiedTrustedCommunity
Run the council workflow from a normal Copilot session only when the user explicitly asks for council review, multi-persona critique, debate, design review, code review, architecture feedback, UX review, or tradeoff analysis. Do not use it for commit, stage, merge, approval, or generic pre-commit requests. Accept the same mode syntax as the bundled council reviewers: `core|auto|core-eng|core-ux|core-mix|all|debate <problem|@file>`. During council slash-command use, the current session agent moderates reviewer agents directly. Runs broader than 6 reviewers require explicit AskUserQuestion approval before launch.
2SKILL.mdUpdated Apr 30, 2026
smykla-skalski/council
tools
VerifiedTrustedCommunity
Use when the user invokes $council, $council:council, Council review, or Council debate. Use loaded SKILL body or one direct installed `skills/council/SKILL.md` read. Direct read path must contain `/.codex/plugins/cache/sai/council/` and end `/skills/council/SKILL.md`. `cd <cwd> && sed -n ... <path>` is valid. Do not use `pwd`, `ls`, `find`, `rg`, `cat`, multiple `&&`, or `;`. Never use repo-local paths. If unavailable, stop exactly `Council not run: skill unavailable.` At most one pre-tool message, exact `Council progress:` line only. Non-final lines start `Council progress:`.
2SKILL.mdUpdated Apr 29, 2026
smykla-skalski/council
development
VerifiedTrustedCommunity
Run council reviews with sourced engineering, UX, reliability, performance, AI, and strategy persona lenses. Use when the user asks for council review, multi-persona critique, debate, design review, code review, architecture feedback, UX review, or tradeoff analysis.
2SKILL.mdUpdated Apr 29, 2026
smykla-skalski/council
development
VerifiedTrustedCommunity
Use when the user asks for council review, multi-persona critique, debate, design review, code review, architecture feedback, UX review, or tradeoff analysis. Bare invocations use `core` profile auto-detect; explicit `auto` selects the best-fit 6 personas from the sourced 27-persona engineering and UX roster. Users can still pin `core`, `core-eng`, `core-ux`, `core-mix`, `all`, or `debate`.
2SKILL.mdUpdated Apr 29, 2026