Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/flashus/detrix/review-plan-autogit clone --depth 1 https://github.com/flashus/detrixWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00014 | $0.01726 |
| Opus 5 | $0.00007 | $0.00863 |
| Sonnet 5 | $0.00003 | $0.00345 |
| Haiku 4.5 | $0.00001 | $0.00173 |
Grade A, and why
review-plan-auto scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are orchestrating an automated plan review-feedback-revision cycle. You act as a project manager coordinating two personas via subagents:
- Reviewer: senior architect + staff developer who finds flaws
- Reviser: the plan author who incorporates feedback
Input
The argument provided is: "$ARGUMENTS"
If the argument is empty or blank, use AskUserQuestion:
Question: "What plan should I review?"
Options:
- "Paste text" — "I'll paste the plan text directly in the chat"
- "Read a file" — "I'll provide a file path to read"
Wait for the user's response before proceeding.
If the argument IS provided:
- If it looks like a file path (contains
/or.md), read the file - Otherwise, treat it as the plan text itself
Setup
Extract a short plan name from the plan's title/first heading (e.g., "auth_fix", "cloud_deploy"). Sanitize to lowercase, underscores, no spaces.
mkdir -p .claude/reviews/<plan_name>
Save the original plan as .claude/reviews/<plan_name>/rev_0.md.
Tell the user: "Starting automated review cycle for plan: <plan_name>. Up to 4 review rounds, stopping early if approved."
Phase 0: Code Verification (before first review)
Launch a subagent to verify the plan's assumptions against the actual codebase:
Prompt: "You are verifying assumptions in a plan against the actual codebase.
Read the plan below, then check every file path, line number, struct name, function name,
and behavioral assumption against the real code. Report:
1. Confirmed assumptions (file exists, line matches, behavior is as described)
2. Wrong assumptions (file moved, line numbers off, struct has different fields, etc.)
3. Missing context (things the plan doesn't mention that are relevant)
Be specific — include actual file:line references from the codebase.
Plan:
<plan text>
Write findings to: .claude/reviews/<plan_name>/verification.md"
Read the verification output. If there are wrong assumptions, they'll be included in the first review.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 225 lines · 14 tokens per session scan A 73acab94529c
review-plan-auto is a command published in the GitHub repository flashus/detrix (23 stars, last pushed 9d ago), licensed MIT. It adds 14 tokens to every session and 1,726 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
freview
Prepare PR with CI checks and Devin AI review loop.
debug-task
Debug Frontman task interactions from the database.
work
Set up a containerized worktree for a GitHub issue or PR.
auto-work
给一个需求，AI自动完成调研→方案→方案Review→开发→开发Review全流程.
develop-review
Review feature-developing 生成的代码，检查遗漏和宪法违规.
developing
Command "developing" from chaohong-ai/ai-auto-work, covering 参数解析, 你的角色, 工作流程, 第一步：建立完整上下文 and 第二步：确认实现范围.