메뉴
Review your radar catalogue in a local web UI — browse, star, dismiss, annotate, and filter discovered tools and insights.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Scan external sources for AI tools, features, and techniques. Builds a discovery catalogue from dependency changelogs, HN (targeted + firehose), GitHub, YouTube, a people-to-watch list, and your inbox.
Match catalogued AI tools and techniques against your goals and usage patterns. Surfaces personalized recommendations.
Your AI development radar — scan the ecosystem for new tools and techniques, then get personalized recommendations. Combines radar-scan and radar-recommend in one command.
Analyze your Claude Code and Cowork session history to surface actionable workflow insights. Diagnoses failures, identifies automation opportunities, aligns time allocation with goals, and flags repeated knowledge worth saving.
| name | radar-review |
| description | Review your radar catalogue in a local web UI — browse, star, dismiss, annotate, and filter discovered tools and insights. |
| argument-hint | (no arguments) |
Launches a tiny local web server and opens the Radar review UI in the user's default browser. The UI lets them star, dismiss, annotate, and filter catalogue items with the full set of filter chips, tier grouping, and score breakdowns.
This skill is self-contained — it does not depend on or use the Open Brain plugin, ai-brain MCP, or any external memory service. All data comes from ~/.claude/radar/catalogue.json and stays local.
Verify ~/.claude/radar/catalogue.json exists. If it does not, tell the user: "No catalogue found. Run /radar-scan first to discover tools and techniques." Stop here.
Run the bundled server in the background:
node "${CLAUDE_PLUGIN_ROOT}/bin/review-server/server.mjs"
Use run_in_background: true. Capture the bash_id so you can read output later.
Read the background output until you see a line starting with RADAR_REVIEW_URL . Extract the URL (everything after that prefix). The server also attempts to auto-open the URL in the default browser.
Tell the user:
Radar review is running at . Browser should have opened automatically — if not, click that URL. Click Done in the UI (or say "done" here) when you're finished.
Wait for the user to either:
/api/exit and the server exits cleanly)Periodically read the background output. When the server exits, it prints a line starting with RADAR_SESSION_SUMMARY followed by a JSON object with session counts.
If the user says "done" but the server is still running, make a quick curl -sS -X POST http://127.0.0.1:<PORT>/api/exit call to trigger graceful shutdown (use the URL extracted in Step 2).
Parse the RADAR_SESSION_SUMMARY JSON line. Output a concise summary:
Review session complete:
- Items starred: N
- Items dismissed: N
- Items reviewed: N
- Items acted on: N
- Notes added: N
Skip lines with zero counts. If no RADAR_SESSION_SUMMARY line was found (e.g., server crashed), just say "Review session ended."