timolins/claude-review
/SKILL.md
Start a code discussion with Claude as a peer reviewer via the CLI. Intended to be called by other AI agents (Codex, Cursor, etc.) after completing their work.
$ install --global
skillsauthnpx skillsauth add timolins/claude-review claude-reviewInstall 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 24, 2026, 8:58 PM1.8s1 file scanned
SKILL.md
- name:
- claude-review
- description:
- Start a code discussion with Claude as a peer reviewer via the CLI. Intended to be called by other AI agents (Codex, Cursor, etc.) after completing their work.
- argument-hint:
- [goal or context for the changes]
- disable-model-invocation:
- true
Peer Code Discussion via Claude CLI
After completing your work, start a code discussion with Claude as a peer. Claude will raise questions and suggestions about your changes. You can push back, disagree, or explain your reasoning — the goal is to reach consensus on the best approach together.
Usage
Start a discussion
~/.agents/skills/claude-review/scripts/review.sh "description of what the changes should accomplish"
Or without a goal (Claude infers from commits):
~/.agents/skills/claude-review/scripts/review.sh
Continue the conversation
Push back, explain your reasoning, or ask Claude to look at updates:
~/.agents/skills/claude-review/scripts/review.sh -c "I did it this way because X, what do you think?"
~/.agents/skills/claude-review/scripts/review.sh -c "I disagree — extracting that would add complexity for no real gain"
~/.agents/skills/claude-review/scripts/review.sh -c "Good point, I've refactored it. How does this look now?"
~/.agents/skills/claude-review/scripts/review.sh -c "I see your concern but this is intentional because of Y"
The -c flag resumes the same session so Claude has full context of the discussion.
Optional tuning (for very large diffs)
You can tune prompt/diff truncation via env vars:
REVIEW_MAX_PROMPT_CHARS=180000 REVIEW_MAX_DIFF_CHARS=120000 REVIEW_MAX_WORKTREE_CHARS=40000 ~/.agents/skills/claude-review/scripts/review.sh
This keeps the review responsive on large branches while still giving Claude enough context.
You can also tune live preview verbosity (stream mode):
REVIEW_LIVE_PREVIEW_MIN_CHARS=160 REVIEW_LIVE_PREVIEW_MIN_EMIT_CHARS=60 REVIEW_LIVE_PREVIEW_INTERVAL_SEC=0 ~/.agents/skills/claude-review/scripts/review.sh
REVIEW_LIVE_PREVIEW_INTERVAL_SEC=0 keeps previews event-driven (newlines/message end).
Set it to >0 only if you want periodic timer-based previews for very long single-line outputs.
Live progress is enabled by default (Claude stream JSON is parsed and shown as it runs).
Set REVIEW_STREAM_JSON=0 to fall back to plain claude --print output:
REVIEW_STREAM_JSON=0 ~/.agents/skills/claude-review/scripts/review.sh
Stream rendering is handled by scripts/render-stream.mjs (Bun first, Node fallback).
If stream parsing fails at runtime, the script auto-retries once in plain output mode so reviews still complete.
The review instruction prompt is stored in scripts/review-prompt.md and loaded by scripts/review.sh, so you can tweak review behavior without editing bash logic.
How it works
- First run: collects branch diff, uncommitted changes, and commit log. Sends to Claude in
--print --output-format=stream-jsonmode (when available), surfaces live progress/tool activity, and saves the session ID to.review-session. - Follow-ups with
-c: resumes that session viaclaude -r <session-id>, includes fresh diff if there are new changes, and continues the discussion.
Important
Claude is set up as a peer, not an authority. It will:
- Raise concerns as questions, not commands
- Concede when you make a good counterargument
- Acknowledge things done well, not just flag problems
The goal is consensus on what's best — not blind compliance with every suggestion.
Related Skills
openclaw/taskflow
tools
VerifiedTrustedCommunity
Use when work should span one or more detached tasks but still behave like one job with a single owner context. TaskFlow is the durable flow substrate under authoring layers like Lobster, ACPX, plugins, or plain code. Keep conditional logic in the caller; use TaskFlow for flow identity, child-task linkage, waiting state, revision-checked mutations, and user-facing emergence.
357,764SKILL.mdUpdated Apr 10, 2026
openclaw/extensions/lobster
tools
VerifiedTrustedCommunity
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
357,764SKILL.mdUpdated Apr 10, 2026
steipete/extensions/lobster
tools
VerifiedTrustedCommunity
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
357,588SKILL.mdUpdated Apr 13, 2026
steipete/xurl
tools
VerifiedTrustedCommunity
A CLI tool for making authenticated requests to the X (Twitter) API. Use this skill when you need to post tweets, reply, quote, search, read posts, manage followers, send DMs, upload media, or interact with any X API v2 endpoint.
356,423SKILL.mdUpdated Apr 13, 2026