A review procedure that checks a specification, plan, code change, or claim against a named project source of truth. It uses a reviewer that did not create the item and returns GO, FIX, or STOP.

Reviewing plans, specifications, diffs, and individual claims before accepting or merging them.

A person or agent can approve its own work without noticing mistakes. Checking against an independent reference makes the review test the work against known requirements.

servo-gate — Skill for Claude Code by t1djani: A review procedure that checks a specification, plan, code…

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 skills/t1djani/servo/servo-gate

Any agent

npx skills add t1djani/servo --skill servo-gate

Clone the repo

git clone --depth 1 https://github.com/t1djani/servo

Made for: Claude Code.

Or install servo, the plugin that ships this one along with the rest of its 9 skills, 1 command.

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/skills/t1djani/servo/servo-gate.svg)](https://agentmods.dev/skills/t1djani/servo/servo-gate)

Your own site

<a href="https://agentmods.dev/skills/t1djani/servo/servo-gate"><img src="https://agentmods.dev/badge/skills/t1djani/servo/servo-gate.svg" alt="Measured on agentmods" height="20"></a>

Per session 57 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 991 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 5.1	$0.00057	$0.00991
Opus 5	$0.00028	$0.00495
Sonnet 5	$0.00011	$0.00198
Haiku 4.5	$0.00006	$0.00099

Measured 6d ago against content hash 4cb394791627, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

servo-gate 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 6d 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.

skills/servo-gate/SKILL.md · 59 lines

How it starts

The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.

servo-gate

Verify an artifact against ground truth, using a non-colluding reviewer: one that did not produce the artifact and judges it only against a named oracle. A gate that does not read an independent oracle is theater — this skill refuses to run without one.

The rule underneath: a check is only worth running if it brings in information the producer did not have. Re-reading the artifact with the same context that wrote it is the agent grading its own homework.

What you need

The artifact: the text under review (a spec, a plan, a diff, a single claim).
The oracle key: which source of truth to check against (e.g. scope, invariants, acceptance). The project's keys live in .servo/manifest.yaml.
The stance (default refute): the posture the reviewer takes.

Procedure

Resolve the oracle. Read .servo/manifest.yaml in the project. Find the key under oracles:. Its value is the source — where the ground truth lives (a file path, git:main, vault:..., linear:...). If the key is not in the manifest, stop and report the available keys; do not guess.
Read the source → the oracle slice. Read the source yourself (Read tool, MCP, git, whatever the source type calls for) and capture the relevant content. This text is the oracle slice. If you cannot read it, stop — never invent ground truth.
Refuse an empty oracle. If the oracle slice is empty or whitespace, stop: a gate with no ground truth is theater. This refusal is by design (§7b).

Build the verifier prompt by filling this exact template:

You are a non-colluding "<STANCE>" reviewer.
You did NOT produce the artifact. Judge it ONLY against the ground truth below.
If the ground truth does not support a claim in the artifact, flag it.

## Ground truth (authoritative oracle)
<ORACLE SLICE>

## Artifact under review
<ARTIFACT>

Reply with JSON only: {"verdict":"GO|FIX|STOP","findings":["..."]}

Dispatch the verifier — the real non-colluding step. Spawn a subagent with the Agent tool and give it exactly the prompt from step 4 and nothing else. It must not see your reasoning or how the artifact was produced. When more than one model is available, run the verifier on a different model than the one that produced the artifact (§7d) to cut shared bias. Tell it to reply with only the JSON object.

Read the full file on GitHub · 59 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.

6d ago First seen · 59 lines · 57 tokens per session scan A 4cb394791627

Subscribe to this mod's changes

servo-gate is a skill published in the GitHub repository t1djani/servo (1 stars, last pushed 2mo ago), licensed MIT. It adds 57 tokens to every session and 991 once invoked, about $0.0003 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 ai-agents claude-code claude-code-plugin claude-plugin claude-skill process

craft

Engineering standards for any code work on any stack — writing production code, slicing and reviewing PRs, authoring tests, running gates, merging, and running agent fleets. Load before touching code, not after review flags it.

abhimanyusingh-gh/software-craftsmanship · 47 tokens

security-architecture-review

Reviews a design or change for security before it ships — authentication and authorization, data handling, secrets, dependencies, and the secure-development practices around it. Use this to review an architecture or pull request for security, set secure coding standards, choose or tune SAST and DAST tooling, assess a…

cbrock84/headcount · 78 tokens

code-review

Conducts and responds to code review — reviewing a change for correctness, design, and risk, and evaluating review feedback received on your own work. Use this before merging, when asked to review a diff or pull request, when review feedback has arrived and needs acting on, or when feedback seems wrong and needs a…

cbrock84/headcount · 72 tokens

servo-gate

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

servo-gate

What you need

Procedure

What this file has done since we first saw it

Other skills, from other repositories

craft

security-architecture-review

code-review

review

pattern-detection

refactor