Spec / Code Drift Check

    1

    Check whether your repository still matches its docs.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Spec / Code Drift Check

    Example session with this skill installed

    Synthetic illustrative example, not a customer session or an Agensi delivery test.

    Compare these two synthetic one-line files only.

    docs/contract.md:1: The client passes a timeout of 5000 milliseconds to the transport.
    src/client.ts:1: return transport.get("/items", { timeoutMs: 3000 });

    Return a static drift report; do not execute code.

    • Read your context and instructions
    • Compiled the spec code drift

    Synthetic worked example

    Scope/method: These two files only; synthetic working-copy snapshot, no verified commit. Static text read: 1/10 document files, 1/20 code files, 1/20 claims, 2/10000 source lines. Complete for the selected claim; no other files inspected. No runtime or transport implementation inspected.

    Summary: 1 Medium finding, 0 High, 0 Low; 0 supported, 0 unresolved.

    Claim ledger

    IDClaimStatusImplementation evidence
    C1docs/contract.md:1: timeout of 5000 millisecondsContradicted (F1)src/client.ts:1: timeoutMs: 3000

    F1: Timeout argument differs [Medium]

    Claim: docs/contract.md:1 says the client passes a timeout of 5000 milliseconds to the transport.
    Code: src/client.ts:1 passes { timeoutMs: 3000 } directly to transport.get.
    Impact: The static request parameter differs from the documented value.
    Confidence: High, based on the direct argument visible in the selected code.
    Uncertainty: Transport enforcement, elapsed timeout and deployed behavior are unverified.
    Next action: Decide the intended timeout and align the claim or argument. Separately test runtime behavior only when authorized.

    Unresolved/next checks: None for this single static claim. Transport enforcement is outside the claim and remains untested. No runtime verification or repository edits were performed.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Compare selected specification or documentation claims with readable repository code, then receive a bounded report with evidence from both sides. Findings include file/line citations, severity, confidence, uncertainty and a suggested next action. Supported and unresolved claims remain visible in a coverage ledger.

    Use your own repository paths and select the documents and code areas to inspect. The host agent must already have authorized file-reading and search access. This instruction-only skill adds no packages, hooks or memory services. It performs static inspection, without executing your code, applying fixes or verifying production behavior. Your host agent and its services handle repository data.

    Derived from my free MIT-licensed Lab-Zero audit workflow: https://github.com/cyb213/Lab-Zero/blob/0c9e21fc65e4a65c9e4e4cbfeb40c48aa50a34f2/.agents/skills/audit/SKILL.md . The original remains freely available. This paid version adds a standalone workflow, explicit coverage ledger, uncertainty handling and an inline worked example.

    This is a small experiment. Small local trials have been done, but reliability and Agensi delivery compatibility are not established. It can miss drift or misread intent: review the cited evidence. No security certification, completeness guarantee or automatic repair is provided.

    May miss drift or misread intent; review cited evidence. Static inspection only: no execution, fixes, production verification, security certification or completeness guarantee. Per-run budgets: 10 document files, 20 code files, 20 claims and 10,000 distinct source lines. Small local trials do not establish reliability or Agensi delivery compatibility. Read-only instructions do not enforce host permissions.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    Permissions

    Read Files

    Reads only user-authorized repository/document paths for static inspection. Returns the report in chat by default. An optional report file requires explicit user authorization and a path outside audited repository roots. Read-only is an instruction, not a permission sandbox or enforcement guarantee. Repository data is handled by the host agent and its services. No additional service accounts are required by this skill.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    Check whether your repository still matches its docs.

    This skill uses the open SKILL.md standard. Use it through an MCP-compatible agent connected to Agensi, and check the compatibility information on this page for confirmed agents and required capabilities.

    Connect your AI agent to Agensi through MCP, then ask it to use Spec / Code Drift Check on your task. Your agent loads the latest approved version each time. Paid skills require a purchase or active subscription. Skill files are not available for download; your licence does not allow you to save, copy or export any part of a skill.

    New in Code Quality & Review