Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Secrets Leak Sentinel

    1

    This skill finds secrets about to be committed — or already in history — and tells you how to contain the damage in the order that actually works.

    $7

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Secrets Leak Sentinel

    Example session with this skill installed

    scan this repo for leaked secretsscan this repo for leaked secrets

    • Read your context and instructions
    • Compiled the secrets leak sentinel

    2 findings (ranked by blast radius):
    🔴 config/prod.env:14 — AWS secret key (AKIA...last2) — LIVE cloud credential.
    🟡 src/client.ts:8 — OpenAI key (sk-p...last2).

    This .env is git-tracked AND already pushed to origin/main.
    Remediation (in order):

    1. ROTATE both keys now — they are public to anyone with the repo.
    2. Remove .env from tracking, add to .gitignore, move values to a secret manager.
    3. Scrub history (git filter-repo) — note: this does NOT un-leak the already-pushed keys; rotation in step 1 is what protects you.masked findings ranked by blast radius + ordered remediation steps for the repo's exact state.

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

    About this skill

    A leaked credential is a breach, not a bug. This skill finds secrets about to be committed — or already in history — and tells you how to contain the damage in the order that actually works.

    Detects: AWS/GCP keys, GitHub/GitLab/Slack/Stripe/OpenAI tokens, private-key blocks, JWTs, and a git-tracked .env / *.pem / service-account JSON — with entropy + placeholder filtering to cut false positives, and allowlist support.

    Reports MASKED. Each finding shows file:line, provider/kind, and a masked preview (first 4 + last 2 chars). The full secret is never printed, stored, or transmitted.

    Remediation in the right order — the valuable part:

    • Staged only: unstage, move to a secret manager, .gitignore.
    • Committed, not pushed: rewrite local history before pushing.
    • Already pushed: ROTATE THE CREDENTIAL FIRST, then scrub history. Rewriting history alone does NOT un-leak an already-public secret.

    Read-only. Never runs a destructive history rewrite or push --force for you.

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

    Does not automate history rewriting or credential rotation. Scans are read-only. Performance may vary on very large git logs.

    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

    Money-back guarantee

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

    Trust & safety

    Security scan

    Security scan passed 18 days ago

    Software
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed18 days ago

    What's inside

    Frequently Asked Questions

    The skill uses provider-specific pattern matching combined with entropy filtering to distinguish between real credentials and harmless mock data, significantly reducing false positive noise compared to basic keyword searches.

    To prevent further exposure, the skill only displays masked previews of detected secrets (e.g., `AKIA...3X7V`) so that raw credentials are never stored in your AI agent's chat history or logs.

    Yes, while it specializes in staged and working tree scans for immediate prevention, you can also request a full historical scan to identify dormant keys hidden in previous commits.

    You receive the core logic for multi-provider detection, risk-based prioritization workflows, and step-by-step remediation guides for rotating and scrubbing leaked data.

    No, the skill is designed to identify risks and provide safe remediation plans; for security reasons, it will not automatically execute destructive commands like `git filter-repo` without your manual oversight.

    It is designed for any Git-based project and supports popular providers including AWS, GCP, GitHub, Stripe, OpenAI, and Slack, while also respecting existing `.gitleaksignore` configurations.

    New in Security