self-review

self-review is a command for coding agents from gaia-react/gaia. It costs 19 tokens per session (1,112 once invoked), scanned A, original, MIT.

A review step for an unfinished software specification before it is shown for approval.

In plain words
What is it for?
It is for checking the draft against an earlier approved snapshot, recording findings, applying low- and medium-priority fixes, and flagging serious issues.
Why use it?
It catches placeholders, unresolved questions, and other problems while the specification is still being drafted.

Command

Install

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.

agentmods
npx agentmods add commands/gaia-react/gaia/self-review
Clone the repo
git clone --depth 1 https://github.com/gaia-react/gaia

Wrote 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.

agentmods badge for self-review

README.md
[![agentmods](https://agentmods.dev/badge/commands/gaia-react/gaia/self-review.svg)](https://agentmods.dev/commands/gaia-react/gaia/self-review)
Your own site
<a href="https://agentmods.dev/commands/gaia-react/gaia/self-review"><img src="https://agentmods.dev/badge/commands/gaia-react/gaia/self-review.svg" alt="Measured on agentmods" height="20"></a>
Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,112 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00019 $0.01112
Opus 5 $0.00010 $0.00556
Sonnet 5 $0.00004 $0.00222
Haiku 4.5 $0.00002 $0.00111

Measured 4d ago against content hash 81c284a0b3d2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

self-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 4d 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.

.specify/extensions/gaia/commands/self-review.md · 90 lines

How it starts

The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Self-review pass before gate 2

The /gaia-spec wrapper dispatches this directly at step 6 as a general-purpose Agent, after the Socratic clarify loop and before the GAIA wrapper presents the rendered artifact for gate-2 confirmation. The extension also declares it on spec-kit's after_clarify event, so a bare /speckit-clarify invocation fires it too.

Inputs

  • The in-progress SPEC draft. The wrapper writes the working draft to .gaia/local/cache/draft.md during the loop; the path may also be supplied via $ARGUMENTS.
  • The gate-1 snapshot at .gaia/local/cache/gate1-<spec_id>.json, captured when the user confirmed intent + UATs.

Audit checklist

Read the draft and the gate-1 snapshot. Record every finding under the headings below (the spec.md 6a dispatch prompt is the authoritative return/apply contract): write them to .gaia/local/cache/audit-<spec_id>/findings/self-review.json, apply the low and medium suggested_fixes to the draft yourself in a single write, and surface the high findings to the wrapper for gating.

1. Placeholder text

Scan for [PLACEHOLDER], <TODO>, <TBD>, FIXME, and bare TBD tokens. Any hit is a finding: name the field and the offending line.

2. Scope drift relative to gate 1

Compare draft fields against the gate-1 snapshot:

  • intent paragraph diverged in shape, not just polish.
  • A UAT was renumbered, removed, or added since gate 1.
  • success_criteria gained or lost an entry without a corresponding answered clarification.

For each drift, name the gate-1 value, the current value, and ask the wrapper to surface a confirmation prompt before gate 2 (or to revert).

3. Internal inconsistency

Cross-check fields against each other:

  • A UAT references a behavior absent from intent or success_criteria.
  • scope_boundaries.always includes a behavior contradicted by scope_boundaries.never.
  • clarifications.answered[] entries that contradict the current intent.

4. Ambiguous UAT phrasing

Read the full file on GitHub · 90 lines

Changes

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.

  1. 4d ago First seen · 90 lines · 19 tokens per session scan A 81c284a0b3d2

Subscribe to this mod's changes

self-review is a command published in the GitHub repository gaia-react/gaia (23 stars, last pushed today), licensed MIT. It adds 19 tokens to every session and 1,112 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.