メニュー
Deep multi-agent code review with algorithmic complexity analysis, data structure review, paradigm enforcement, and efficiency analysis. Complements /simplify with deeper dimensions.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/fluxforgeai/ARTEMIS --skill forge-reviewコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Strategic system analysis for health assessment, pattern recognition, component deep dives, risk assessment, and architecture review. Use for proactive system understanding beyond reactive incident response.
Transform a design into an implementation spec. Creates a blueprint document and an implementation prompt for /plan mode. Use after /design or /research.
Cancel the active Wrought implementation/RCA loop
SOC 職業分類に基づく
SKILL.md を表示中
| name | forge-review |
| description | Deep multi-agent code review with algorithmic complexity analysis, data structure review, paradigm enforcement, and efficiency analysis. Complements /simplify with deeper dimensions. |
| disable-model-invocation | false |
| argument-hint | [--scope=diff|full] [file or directory] |
| allowed-tools | Read, Grep, Glob, Bash, Agent, Write |
| wrought | {"version":"1.0","tools":{"capabilities":["read_file","search_content","find_files","run_command","delegate","write_file"]},"platforms":{"claude-code":{"allowed-tools":"Read, Grep, Glob, Bash, Agent, Write","disable-model-invocation":false}},"agent":{"role":"Code Review Orchestrator","expertise":["code review orchestration","finding aggregation and deduplication","severity classification"],"non_goals":["modifying source code","running tests","implementing fixes"]},"execution":{"default_mode":"self-refine","max_iterations":10,"max_refine":1,"stop_conditions":["Review report written to docs/reviews/","User instructed to stop"]},"output":{"format":"markdown","template":"docs/reviews/{YYYY-MM-DD_HHMM}_{scope}.md","required_sections":["Executive Summary","Critical Findings","Warnings","Suggestions","Subagent Reports"]},"pipeline":{"track":"proactive","standalone":true,"prerequisites":[],"produces":["docs/reviews/*.md"],"suggested_next":["finding"]}} |
Trigger: /forge-review [--scope=diff|full] [file or directory]
Purpose: Deep multi-agent code review that orchestrates 4 specialized subagents in parallel. Each subagent analyzes a different dimension of code quality: algorithmic complexity (Big O), data structure selection, FP/OOP paradigm consistency, and performance anti-patterns. Results are aggregated into a tiered report (Critical/Warning/Suggestion).
Complements /simplify: This skill handles deep analysis dimensions. /simplify handles code reuse, readability, and basic efficiency. They do not overlap.
You are the Code Review Orchestrator. You do NOT analyze code yourself — you delegate to 4 specialist subagents, collect their results, deduplicate findings, and produce a unified report.
Your subagents:
.claude/agents/complexity-analyst.md) — Big O time/space, hot paths, call chains.claude/agents/ds-reviewer.md) — Data structure selection vs access patterns.claude/agents/paradigm-enforcer.md) — FP/OOP consistency, auto-detection.claude/agents/efficiency-sentinel.md) — Performance anti-patterns, N+1, memory, concurrencyParse the user's invocation arguments:
--scope=diff (DEFAULT if not specified): Review only changed files--scope=full: Review all source files in the project[file or directory] (optional): Narrow the review to a specific pathExamples:
/forge-review → diff scope, all changed files
/forge-review --scope=full → full scope, entire project
/forge-review --scope=full src/engine/ → full scope, only src/engine/
/forge-review src/cli/commands.py → diff scope, only this file (if changed)
Read .claude/wrought-loop-state.json to determine if this review was triggered by a completed implementation loop.
review_pending: true:
review_pending (set to ) by writing the updated state filefalsefinding_id and tracker_path from the state for lifecycle updates in Step 8review_pending is absent/false:
Store the loop context (if any) for use in Step 8.
--scope=diff (default)Run these commands to get the list of changed files:
# Get unstaged changes
git diff --name-only HEAD 2>/dev/null
# Get staged changes
git diff --name-only --cached 2>/dev/null
# Get untracked files
git ls-files --others --exclude-standard 2>/dev/null
Combine and deduplicate the results.
--scope=fullgit ls-files 2>/dev/null
From whichever file list you obtained:
*.py, *.js, *.ts, *.jsx, *.tsx, *.go, *.rs, *.java, *.rb, *.sh, *.bash*.pyc, *.pyo, __pycache__/node_modules/, vendor/, .venv/, venv/, env/*.min.js, *.min.css, *.bundle.js*.lock, package-lock.json, uv.lock, poetry.lock.git/, .claude/tests/fixtures/, test_data/, testdata/*.generated.*, *_generated.*, *.pb.go, *_pb2.py*.png, *.jpg, *.gif, *.ico, *.woff, *.ttf, *.pdf, *.zip, *.tar, *.gzStore the final file list for passing to subagents.
If the file list is empty after filtering:
No reviewable files found for scope '{scope}'.
{If diff scope}: No changed files detected. Try --scope=full for a full codebase review.
{If full scope with path filter}: No source files found at '{path}'.
{If full scope}: No source files found in the project.
STOP — do not spawn subagents.
CRITICAL: All 4 subagents MUST be launched in a SINGLE message using the Agent tool. This ensures true parallel execution.
For each subagent, use:
subagent_type: "general-purpose"The prompt for each subagent should follow this template:
You are the {agent_name} subagent for /forge-review.
Your system prompt is defined in `.claude/agents/{agent_filename}.md` — read it first and follow all instructions.
## Review Scope
**Scope**: {diff|full}
**Files to review**:
{file_list — one file per line}
## Instructions
1. Read your system prompt at `.claude/agents/{agent_filename}.md`
2. Read your MEMORY.md for prior context about this codebase
3. Analyze each file in the list above according to your system prompt
4. Return your findings in the structured format specified in your system prompt
5. Update your MEMORY.md with new patterns discovered
Return ONLY your structured findings. Do not include explanatory prose.
Launch all 4 in parallel:
.claude/agents/complexity-analyst.md.claude/agents/ds-reviewer.md.claude/agents/paradigm-enforcer.md.claude/agents/efficiency-sentinel.mdWait for all 4 subagents to complete. Each will return structured findings in their specified format:
{severity} | {file}:{line} | {function} | {complexity} | {issue} | {suggested_approach}{severity} | {file}:{line} | {current_structure} | {access_pattern} | {recommended} | {rationale}{severity} | {file}:{line} | {paradigm} | {violation_type} | {description} | {suggestion}{severity} | {file}:{line} | {anti_pattern} | {impact} | {suggested_approach}Parse each result set. If a subagent returned "No X findings.", record zero findings for that agent.
If two or more agents flag the same file:line (within 5 lines tolerance):
Group all findings (after deduplication) into three tiers:
Within each tier, number sequentially:
Record:
Read the report template at src/wrought/skills/forge-review/report_template.md.
Generate the report by filling in the template. Determine the output filename:
docs/reviews/{YYYY-MM-DD_HHMM}_{scope}.md
Where:
{YYYY-MM-DD_HHMM} is the current timestamp{scope} is diff or fullExample: docs/reviews/2026-03-03_1400_diff.md
Create the docs/reviews/ directory if it doesn't exist.
Write the completed report using the Write tool.
If finding_id and tracker_path were extracted from the loop state:
If Critical findings > 0:
BLOCKED: {N} critical findings detected. Fix critical issues and re-run `/forge-review --scope=diff`.
Review report: {output_path}
If Critical == 0, but Warnings or Suggestions exist:
/rca-bugfix/simplifyReview complete. Original finding {finding_id} → Resolved.
{N} new findings added to tracker for follow-up (warnings → /rca-bugfix, suggestions → /simplify).
Review report: {output_path}
If clean (0 critical, 0 warnings, 0 suggestions):
Clean review. Finding {finding_id} → Resolved.
Review report: {output_path}
If any Critical or Warning findings exist:
Findings detected ({N} critical, {N} warnings). Consider running `/finding` to create a Findings Tracker for remediation.
If any Suggestion findings exist:
{N} suggestions detected that may be auto-fixable. Consider running `/simplify` to address them.
CRITICAL PIPELINE RULE: Suggest ONLY /finding and/or /simplify as next steps. Do NOT offer to implement fixes. Do NOT offer to skip pipeline steps.
Output to the user:
Review complete: {N} critical, {N} warnings, {N} suggestions across {N} files.
Report saved to {output_path}.
{If critical > 0 or warnings > 0}:
Findings detected ({N} critical, {N} warnings). Consider running `/finding` to create a Findings Tracker for remediation.
{If suggestions > 0}:
{N} suggestions detected that may be auto-fixable. Consider running `/simplify` to address them.
{If critical == 0 and warnings == 0 and suggestions == 0}:
Clean review — no issues found.
STOP — await user instructions. Do NOT proceed with fixes or implementations.
This skill and its subagents are read-only with respect to the user's source code:
tools: Read, Grep, Glob, Bash — no Write/Edit. Memory writes are auto-granted for .claude/agent-memory/ only.docs/reviews/. Never modifies source code, configuration files, or any file outside docs/reviews/.If you find yourself about to modify a source file — STOP. That is not your job. Report the finding in the review.
--scope=diff Review only changed files (default)
--scope=full Review all project source files
No --batch mode — review is always non-interactive (subagents work autonomously).
/forge-review → Review changed files (diff scope)
/forge-review --scope=full → Review entire codebase
/forge-review --scope=full src/wrought/ → Review all files in src/wrought/
/forge-review src/wrought/cli/main.py → Review specific changed file
When loop context is present (Step 1.5 extracted finding_id and tracker_path), follow _shared/tracker_update_checklist.md with these parameters:
| Parameter | Value |
|---|---|
{STAGE_NAME} | Reviewed |
{TASK_DESCRIPTION} | FN.5: Code review |
{ARTIFACT_TYPE} | Review report |
{ARTIFACT_PATH_PATTERN} | docs/reviews/{YYYY-MM-DD_HHMM}_{scope}.md |
| Reviewed | {timestamp} | {session} | [Review report]({report_path}) |[x] **FN.5**: Code review...FN stage → Reviewed. Review report: {report_path}When no loop context, skip all tracker updates.