Borrowing it
Nothing to install: this file belongs to WingedGuardian/GENesis-AGI. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/WingedGuardian/GENesis-AGI/main/.claude/commands/deep-review.mdgit clone --depth 1 https://github.com/WingedGuardian/GENesis-AGIWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/wingedguardian/genesis-agi/deep-review)<a href="https://agentmods.dev/commands/wingedguardian/genesis-agi/deep-review"><img src="https://agentmods.dev/badge/commands/wingedguardian/genesis-agi/deep-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/wingedguardian/genesis-agi/deep-review"><img src="https://agentmods.dev/badge/commands/wingedguardian/genesis-agi/deep-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00000 | $0.02653 |
| Opus 5.5 | $0.00000 | $0.01061 |
| Sonnet 5.5 | $0.00000 | $0.00531 |
| Haiku 4.5 | $0.00000 | $0.00265 |
Grade A, and why
deep-review 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 3d 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 — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Run a fresh-context, un-anchored ADVERSARIAL-LOGIC review of the current diff before pushing — the review that catches real bugs, not a lint/secrets/"structural CLEAN" scan. Use it on any substantial change, and ALWAYS before pushing a hook, gate, parser, or security-surface change.
Why this exists: an internal review-loop analysis of recent PRs found that first-round Codex findings were real, pre-existing bugs — the initial pushes had been self-reviewed with the WRONG SHAPE (a pattern/secrets scan), not an adversarial logic audit. This command makes the right-shape pass the easy default. (It reliably kills the first round; it is NOT a cure for the whole loop — fix-churn and hand-rolled parsing drive the tail. See build-session's Common Traps entry on CLI parsing.)
Note: /deep-review is the LOCAL pre-push adversarial pass (Claude-model reviewers). It does
NOT replace the independent-model Codex review, which runs on the PR — automatically when the PR
opens, and after any later push only when you comment @codex review (an
owner-tunable setting that has flipped before — verify at the PR rather than
trusting this sentence) — and is required by the
merge gate. The two are complementary (Codex catches cross-model blind spots). This command
clears the local commit review-depth gate; it does not certify the PR by itself.
A build session's work on an item ends when the PR EXISTS — not here. Clearing this
gate merely unblocks the commit that this command's own later steps make, and the push and
gh pr create after it: stopping at the gate would end the session before the durable
handoff exists, leaving finished work in a local worktree that no closing session can see
(Codex P2, PR #1638). From the open PR onward — the Codex round, the findings, the merge —
the item belongs to a closing session (.claude/skills/closing-session/), which is why
this command does not tell you to wait for a review it never triggers.
1. Stage everything, then scope the FULL branch diff
ROOT="$(git rev-parse --show-toplevel)"— the repo root. Use$ROOT/scripts/…for everyscripts/…path below: this repo's Bash cwd drifts, so a bare relativescripts/…can resolve against the wrong directory and fail. (gitcommands find the root on their own.)- Stage all intended changes first (
git add …) and skimgit status:git diffnever shows UNTRACKED (never-added) files, so a brand-new source/test/hook/config file is invisible to the review until it is staged. Nothing untracked = the diff is complete. git diff "$(git merge-base HEAD origin/main)"— everything the branch changes vs the merge-base, INCLUDING staged and unstaged tracked work (git diff <commit>compares the working tree). Add--statfor the file list. Do NOT use..HEADalone (misses uncommitted work) or plaingit diffalone (misses STAGED work).
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.
- 3d ago Changed cec25c46e39c
- 12d ago Changed · +3 lines c42dd706fd51
- 1mo ago Changed · +7 lines 1b4471e881c0
- 1mo ago Changed · +8 lines de18c56c0135
- 1mo ago Changed · +4 lines d1ce9290deda
- 1mo ago Changed · +3 lines f22287d23ede
- 1mo ago Changed · +39 lines 2a6006c0c7a3
- 1mo ago Changed · +3 lines f2876b1b9126
- 1mo ago First seen · 84 lines · 0 tokens per session scan A 175bb64e5d58
deep-review is a command published in the GitHub repository WingedGuardian/GENesis-AGI (99 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,653 tokens. 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
review-prs
Check open PRs for human feedback before picking new work.
brief-refresh
Audit docs for staleness and contradictions, propose new features.
refactor-clean
Refactor provided code for cleanliness, maintainability, and alignment with SOLID principles and modern best practices — no over-engineering.
ingest-cwe-taxonomies
Ingest CWE XML from MITRE and generate per-domain security audit taxonomies for PromptKit.
review
A command for reviewing code with security checks and a check for compliance with MX tags. The available description does not explain what MX tags are.
default-ai-review
Extract findings from the provided diff. Rank by severity. Produce a structured review. Emit one recommendation.