A command that compares an implementation with its authoritative specification, meaning the document that defines what the implementation must do. It classifies each requirement and provides evidence and a remediation plan.

Use it to audit code against a specification, verify requirements and interface contracts, and plan fixes for discrepancies.

It helps identify missing, incomplete, conflicting, or changed behaviour before the implementation is treated as correct.

vet-impl — Command for Codex by sergeyklay: A command that compares an implementation with its authoritative…

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/sergeyklay/.agents/vet-impl

Clone the repo

git clone --depth 1 https://github.com/sergeyklay/.agents

Made for: Codex.

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.

README.md

[![agentmods](https://agentmods.dev/badge/commands/sergeyklay/.agents/vet-impl.svg)](https://agentmods.dev/commands/sergeyklay/.agents/vet-impl)

Your own site

<a href="https://agentmods.dev/commands/sergeyklay/.agents/vet-impl"><img src="https://agentmods.dev/badge/commands/sergeyklay/.agents/vet-impl.svg" alt="Measured on agentmods" height="20"></a>

Per session 10 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 614 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.

Model	Per session	Once invoked
Fable 5	$0.00010	$0.00614
Opus 5	$0.00005	$0.00307
Sonnet 5	$0.00002	$0.00123
Haiku 4.5	$0.00001	$0.00061

Measured 5d ago against content hash 634cfc74c60c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

vet-impl 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 5d 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.

.agents/commands/vet-impl.md · 33 lines

How it starts

The opening of the file, as written. The whole thing — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Your task is to verify that an implementation faithfully realizes every requirement, constraint, design decision, interface contract, algorithm, and invariant defined in its authoritative specification.

Task

Use the verify-impl Agent Skill to drive every phase: build project context, extract every verifiable requirement, discover the implementation surface, classify each requirement (PASS / DRIFT / PARTIAL / MISSING / CONFLICT) with evidence, run the cross-cutting and self-verification passes, then emit the verdict and remediation plan.
Treat the user-provided argument as the path to the specification file. Read it in its entirety before classifying anything.
Discover project context by searching the conventional locations - project context files (AGENTS.md, CLAUDE.md, GEMINI.md, CONTRIBUTING.md, README.md), architecture documentation, ADR sets, build manifests, and schema files. Do not assume specific filenames or paths; the skill's discovery rules apply.

Skill Enforcement

MANDATORY: Apply the verify-impl Agent Skill verbatim.

The skill is the single source of truth for how to extract requirements, classify findings, and structure the report. The project's context files are the source of truth for what the architecture, layering, naming conventions, and review standards are. Consult each at the moment the skill calls for it.

Process:

Load the verify-impl Agent Skill before reading the specification.
Walk through every phase in order. Do not skip, merge, or abbreviate any phase. Each gate prevents a documented failure mode (recency bias, false-positive amplification, severity inflation, missed cross-cutting concerns).
If the skill is unavailable in this environment, stop and report the failure. Do not improvise a replacement protocol.
Write the final review to .reviews/Review-{spec-name}.md per the skill's output rules. Print the path of the file written.

Constraints

Review only. Do not write, modify, refactor, or restructure any source code. A finding proposes a change; an implementer applies it later. Do not commit, branch, or open pull requests.
Quote before judging. Extract the exact spec text before evaluating code against it. This prevents drift between what you think the spec says and what it actually says.
Absence is a finding. Behavior defined in the spec but not implemented is MISSING, not "probably handled elsewhere".
The spec is the authority. A divergent implementation that "seems better" is still a defect. If the spec itself is wrong, raise it as a separate flag - do not excuse the implementation divergence.
No severity inflation. Reserve critical for safety violations, data loss, and behavioral contradictions. Reserve major for correctness bugs and missing functionality. Reserve minor for naming and style divergence.

Read the full file on GitHub · 33 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.

5d ago First seen · 33 lines · 10 tokens per session scan A 634cfc74c60c

Subscribe to this mod's changes

vet-impl is a command published in the GitHub repository sergeyklay/.agents (5 stars, last pushed today), licensed Apache-2.0. It adds 10 tokens to every session and 614 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 2026-08-31.

Code review agent-skills agentic-workflow agents

test-feature

Test a React Native feature on the running simulator/emulator. Verifies UI, user flows, and internal state. Generates a persistent Maestro test file.

Lykhoyda/rn-dev-agent · 30 tokens

nav-graph

Extract, inspect, and query the app navigation graph — a complete map of all screens and navigators.

Lykhoyda/rn-dev-agent · 21 tokens

doctor

Diagnose installation health. Check Node, CDP bridge, rn-fast-runner (iOS), rn-android-runner (Android), maestro-runner, simulators, Metro, CDP, injected helpers, ffmpeg, physical devices, plugin version, Vercel rules sync. Reports what's missing — does NOT modify your project.

Lykhoyda/rn-dev-agent · 68 tokens

vet-impl