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.
git clone --depth 1 https://github.com/andreidavid/codex-reviewWrote 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/andreidavid/codex-review/codex-review-waive)<a href="https://agentmods.dev/commands/andreidavid/codex-review/codex-review-waive"><img src="https://agentmods.dev/badge/commands/andreidavid/codex-review/codex-review-waive/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/andreidavid/codex-review/codex-review-waive"><img src="https://agentmods.dev/badge/commands/andreidavid/codex-review/codex-review-waive.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 | $0.00013 | $0.00826 |
| Opus 5.5 | $0.00005 | $0.00330 |
| Sonnet 5.5 | $0.00003 | $0.00165 |
| Haiku 4.5 | $0.00001 | $0.00083 |
Grade A, and why
codex-review-waive 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 1mo 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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Waive a Codex Review Finding
The user invoked this command with: $ARGUMENTS
Waivers are per-repo suppressions for review findings the user has decided to
accept. Waived findings no longer block the post-commit hook; they still
appear in the review history, marked "waived": true.
Steps
-
Resolve the git dir and history file:
GITDIR=$(git rev-parse --git-common-dir)History:
$GITDIR/codex-reviews.jsonl. Waive file:$GITDIR/codex-review-waived. -
Read the most recent FAIL entry for the current commit. History is repo-wide and another session may have failed more recently, so resolve your own failure via your worktree's HEAD — never take the latest FAIL alone:
WAIVED_SHA=$(git rev-parse HEAD) jq -s --arg sha "$WAIVED_SHA" \ '[.[] | select(.verdict=="FAIL" and .sha==$sha)] | last' \ "$GITDIR/codex-reviews.jsonl"If there is none (or the file is missing), tell the user there is no failed review of the current commit (
HEAD) to waive findings from, and stop. -
List its blocking findings (priority
P1/P2withwaivedfalse), numbered, showing priority and title. Select per$ARGUMENTS:- a number → that finding
all→ every blocking finding- a text snippet → the finding whose title matches it unambiguously
- no arguments and exactly one blocking finding → that finding
- otherwise → ask the user which finding(s) to waive
-
For each selected finding, append its
waive_keyfield VERBATIM as one line to$GITDIR/codex-review-waived(create the file if needed; skip keys already present). Never derive the key from the title yourself — always copy thewaive_keyvalue from the history entry, since the hook computes it with a specific normalization. -
If every blocking finding in that entry is now waived, release the block — but only for the review you just waived. State files are per-session (
.<session-id>suffixes) and other sessions may have unrelated failing reviews that must stay blocked, so remove only state whose recorded sha matches$WAIVED_SHAfrom step 2, plus its loop counter:for f in "$GITDIR"/codex-review-state*; do [ -f "$f" ] || continue read -r _ state_sha _ _ < "$f" if [ "$state_sha" = "$WAIVED_SHA" ]; then rm -f "$f" "$GITDIR/codex-review-loop-count${f#"$GITDIR"/codex-review-state}" fi done
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.
- 1mo ago First seen · 76 lines · 13 tokens per session scan A 3bf19f695b78
codex-review-waive is a command published in the GitHub repository andreidavid/codex-review (3 stars, last pushed 2mo ago), licensed MIT. It adds 13 tokens to every session and 826 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-08-31.
Other commands, from other repositories
git
The pre-finish status: branch, hygiene findings, message checks, workflow lint, template state.
ship
Branch, commit, open PR, gather Claude + every enabled AI reviewer (Copilot, CodeRabbit, etc.), fix/justify/resolve every finding, loop until clean, then merge. Run only when implementation is finished AND the owner has said to ship (e.g. "ship it") — never self-invoke just because the work looks done. To design and…
gate
The consensus gate — opt this repo in (or out) of duet-signed commits.
diff-review
Generate a visual HTML diff review — before/after architecture comparison with code review analysis.
review
Dispatches the independent turnstile-reviewer agent to verify a ticket is actually done. On pass, commits the change. Gate 2 of the turnstile workflow -- the implementer never self-certifies.
fire-and-forget
"fire and forget" / "fire and forget": a raw ask or sizing to a reviewed commit in one fire.