More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    SpecProof — Test Integrity & Mutation Gate

    1

    Deterministic AST test integrity, diff-scoped micro-mutation (TS/JS, Python, Go), and SPEC.md coverage gate via MCP, CLI, and GitHub Actions.

    Free

    Security scanned
    SpecProof — Test Integrity & Mutation Gate

    SpecProof — Test Integrity & Mutation Gate

    Example session with this skill installed

    Run a full SpecProof check on my current changes against origin/main. Enable mutation testing with a 75% failure threshold and verify against SPEC.md.

    • Read your context and instructions
    • Compiled the specproof

    I have initiated the SpecProof verification suite on the current diff against origin/main. The process included AST integrity scanning, diff-scoped micro-mutation, and specification mapping.

    SpecProof Verification Report

    CategoryStatusDetails
    Test IntegrityPASS12 tests scanned. 0 vacuous assertions. 0 skipped tests.
    Mutation TestingFAIL72% killed (Threshold: 75%). 3 mutants survived in src/auth.ts.
    Spec MappingPARTIAL4/5 requirements met. Requirement [REQ-05] has no matching test coverage.

    Integrity Findings

    • Vacuous Assertions: None found.
    • Mocking: src/auth.test.ts shows balanced mock-to-assertion ratio (2:5).

    Mutation Results

    Survived: src/auth.ts:42 - Mutated >= to >. The test suite passed, indicating a missing boundary test for the exact limit value.

    • Survived: src/auth.ts:88 - Mutated || to &&. Logic path for secondary auth provider is not asserted.
    • Timeouts: 0 (No infinite loop risks detected).

    Spec Coverage Gap

    • Requirement: [REQ-05] System must lock account after 5 failed attempts.

    Status: Code changes found in src/auth.ts, but no corresponding assertions found in src/auth.test.ts or src/integration.test.ts.

    Next steps

    • Add a boundary test case to src/auth.test.ts specifically checking the equality condition on line 42.
    • Implement a test for the account lockout logic to satisfy [REQ-05].
    • Re-run run_diff_mutation to verify the new tests kill the surviving mutants.

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

    About this skill

    Stop AI coding agents from passing CI with vacuous assertions (expect(true).toBe(true)), empty or skipped tests, swallowed errors, and excessive mocking.

    SpecProof replaces guesswork with execution-backed proof:

    • Zero-LLM AST Test Integrity Gate (TS/JS, Python & Go): Detects vacuous assertions, empty/skipped tests, mock abuse, and swallowed catch/except blocks.
    • Diff-Scoped Micro-Mutation Engine: Mutates only changed lines inside isolated sandboxes and runs your real test suite in seconds to catch missing edge-case checks (e.g., >= vs >).
    • Spec-to-PR Coverage Mapping: Verifies SPEC.md or inline acceptance criteria against changed symbols and tests, with optional BYOK OpenAI / Anthropic / Gemini evaluation.
    • MCP + CLI + GitHub Actions: Exposes check_test_integrity, run_diff_mutation, and verify_spec_coverage for Cursor, Claude Code, and Antigravity.

    Open-source core (Apache-2.0): https://github.com/Claw2M/specproof

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    Requires Node.js 22.13.0+ and git. Mutation testing requires a functional local test command. Supports TS/JS, Python, and Go only.

    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 buying a skill.

    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

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed today

    • Free with an Agensi account

    Software
    Node.js@specproof/cliGit
    Optional
    GitHub CLI / Token
    OpenAI
    orAnthropic
    orGemini API key (BYOK)
    Docker

    Source & quickstart: https://github.com/Claw2M/specproof (Node.js =22.13.0). Runs deterministic AST & isolated sandbox mutation checks locally; network is only used when BYOK LLM (--llm), GitHub PR comments, or Team Cloud (--cloud) flags are enabled.

    Permissions

    Terminal / Shell
    Read Files
    Network Access
    Write Files

    File Scopes

    src/**
    test/**
    .specproof.json
    SPEC.md
    Listedtoday

    What's inside

    Frequently Asked Questions

    SpecProof uses deterministic AST analysis and diff-scoped micro-mutation to verify that tests actually fail when logic changes, whereas standard tools only check if lines were executed, regardless of whether assertions were valid.

    It is compatible with any MCP-enabled AI agent, such as Cursor or Claude Code, and can also be integrated directly into your CI/CD pipeline using the CLI.

    No, the 'Diff-Scoped' approach only mutates code changed in your current branch or PR, allowing for fast, targeted feedback without the overhead of testing the entire codebase.

    SpecProof supports TypeScript, JavaScript, Python, and Go, requiring Node.js 22.13.0 or higher and a Git-initialized project to analyze diffs.

    The Spec-to-PR Mapping feature scans your markdown documentation (like SPEC.md) and correlates your defined acceptance criteria with the actual code and tests present in your pull request.

    You receive the full SpecProof skill package, including the MCP server, CLI tool, and documentation for integrating the integrity and mutation gates into your workflow.

    New in LLM Ops

    Free