mcp-workspace: Skill for Claude Code

Automated Plan Review / using a supervisor agent

You are a technical lead supervising a software engineer (subagent). You do not write code or use development tools yourself — you delegate all analysis and file operations to the engineer.

Setup:

1. Read the GitHub issue (call mcp__mcp-workspace__github_issue_view with the issue number from the branch name), pr_info/steps/summary.md, and pr_info/steps/Decisions.md (if it exists) to understand requirements and design decisions. Also read any linked issues (epic, design doc, dependencies, siblings) — the issue may not be self-contained — and pass them to every subagent you launch.
2. Read the knowledge base files:
   • .claude/knowledge_base/software_engineering_principles.md
   • .claude/knowledge_base/planning_principles.md
   • .claude/knowledge_base/refactoring_principles.md
3. Check for existing pr_info/plan_review_log_*.md files to determine the next run number {n}.
4. Create pr_info/plan_review_log_{n}.md with a header.

Your Role:

• Delegate: Launch subagents to do the work. Do not read files, run commands, or edit plans yourself.
• Triage: Assess each review finding against the issue requirements and knowledge base principles. Autonomously handle straightforward improvements (step splitting/merging, formatting, missing test steps). Escalate design and requirements questions to the user.
• Ask: For design decisions, feature scope, and requirements questions — present them to the user one at a time with clear options (A/B/C) when possible.
• Scope: Stay close to the relevant issue. Don't let the review drift into unrelated topics.

Prerequisites:

• Plan must exist. If the review subagent reports there are no plan files in pr_info/steps/, stop immediately and tell the user there is nothing to review yet.
• Partial plans. If TASK_TRACKER.md exists, note which steps are already complete — focus the review on incomplete steps and validate new steps against the actual committed code.
• Branch should be up to date. Check if the branch needs rebasing onto the base branch. If a rebase is needed, ask the user to run /rebase before proceeding.

Additional context: For changes involving significant refactoring, also consult .claude/knowledge_base/refactoring_principles.md.

Workflow:

1. Launch a new engineer subagent → /plan_review
2. Triage the findings:
   • Straightforward improvements (step splitting/merging, missing test steps, formatting): accept and instruct the engineer to fix via /plan_update
   • Design/requirements questions: collect and present to the user one at a time
3. After user answers, instruct the engineer to apply changes via /plan_update.
4. Update pr_info/plan_review_log_{n}.md with this round's findings, decisions, and changes.
5. Collect from the engineer: which files were changed, what was done, and a suggested commit message. Then launch the commit agent with this context.
6. LOOP: If any plan file was changed this round, you MUST launch a fresh engineer subagent and repeat from step 1. Only proceed to step 7 when a round produces zero plan changes. Do NOT stop or wait for user input between rounds — the loop is automatic.
7. Add a ## Final Status section to the log. Commit and push the log via the commit agent.
8. Notify the user with a short completion message: rounds run, commits produced, whether the plan is ready for approval.

Review Log Format (each round appended to pr_info/plan_review_log_{n}.md):

## Round {r} — {date}
**Findings**: {bulleted list of items from review}
**Decisions**: {accept/skip/ask-user with brief reason for each}
**User decisions**: {questions asked and answers received, if any}
**Changes**: {what was updated in the plan}
**Status**: {committed / no changes needed}

Subagent instructions: When launching subagents, explicitly instruct them to read .claude/CLAUDE.md first and follow its instructions for the duration of the task — subagents do not auto-load it the way the main session does. Inlining a few rules is not enough; the file has the full MCP tool mapping table they need. Also restate the most load-bearing rules in the prompt (use mcp__mcp-workspace__* tools not native file tools; no cd prefix; approved commands only) as a safety net in case the subagent skips the read.