More screenshots
Works with the AI Agents you already use
trishula secret sentinel
Offline, zero-dependency secret scanner to detect leaked API keys, tokens, and private keys in code and git history.
$5
/moSubscription · cancel anytime
Secure checkout via Stripe
- Always the latest version
- Delivered live to your agent via MCP
- Cancel anytime, access ends at period end
trishula secret sentinel
Example session with this skill installed
Scan my staged changes for secrets before I commit this. Use text format and fail if anything critical is found.
- Read your context and instructions
- Compiled the trishula secret sentinel
I have performed a scan of the staged changes in the git index. Two critical findings were detected in your configuration files. You should rotate these credentials immediately before proceeding.
Secret Sentinel Scan Results
SECRET SENTINEL v1.1.0 | verdict: FINDINGS_DETECTED
src/auth/provider.py:12: [CRITICAL] OPENAI_API_KEY - OpenAI API Key sk-************************************************
infra/k8s/secrets.yaml:45: [CRITICAL] DB_URL - Database Connection URL postgres://admin:****************@db.internal:5432/prod
files: 2 | findings: 2 (CRITICAL 2, HIGH 0, MEDIUM 0, LOW 0)
attestation sha256: 4f7e9b2c1a3d5f6e8d9c0b1a2f3e4d5c6b7a890123456789abcdef0123456789
Next steps
1.
Rotate credentials: Immediately invalidate the OpenAI key and the database password through their respective provider dashboards.
2. Remove secrets: Replace the hardcoded values with environment variable references or a secure secret manager.
3. Re-scan: Run the staged scan again after updating the files to ensure no other leaks are present.
Subscription · always the latest version
- Secure checkout via Stripe
- Cancel anytime
- Delivered live via MCP
About this skill
AI coding agents often accidentally commit or share sensitive credentials found in environment files, configuration, or hardcoded strings. This skill provides an offline, zero-dependency scanner to detect leaked cloud keys, LLM provider tokens, and database URLs before they leave the local machine.
What it does
- Multi-source scanning detects secrets in the current working tree, staged git changes, or full repository history.
- Pattern matching identifies 34 specific credential shapes including AWS, GitHub, OpenAI, Anthropic, Stripe, and private key blocks.
- Privacy-first reporting redacts all matches in the output so the agent never sees or logs the actual secret.
- Entropy analysis flags high-entropy strings that may be obfuscated keys or unique identifiers.
- Tamper detection generates SHA-256 attestations to verify if files have changed since the last clean scan.
How it works
- Working tree scan runs against the entire project directory to find existing vulnerabilities.
- Pre-commit check scans only the lines added to the git staging area to prevent new leaks.
- History audit searches through past commits to find secrets that were deleted but remain in the git log.
- Verification uses the attestation command to ensure previous scan results are still valid for the current file states.
Frameworks & tools
This tool is built with standard-library Python 3.10+, requiring no third-party packages. It integrates with Git for history and staging analysis.
Why this beats prompting it yourself
Generic prompts often fail to recognize specific regex patterns for obscure providers and may inadvertently cause the LLM to output the secret in plain text. This skill enforces strict redaction and uses a dedicated scanning engine that runs locally without sending code to an external API.
Use cases
- Prevent committing
.envfiles or hardcoded LLM keys during an autonomous coding session. - Audit a legacy repository for leaked database connection strings or RSA private keys.
- Generate a SARIF report for local CI/CD verification before pushing a PR.
- Verify that a directory has not been tampered with since its last security scan.
Known limitations
Limited to 34 patterns compared to larger tools like Gitleaks. Does not verify credential validity via network. Files over 1MB or containing NUL bytes are skipped. No ignore file support.
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.
34 patterns only. No network-based validity checks. No .gitignore-style ignore file. Skips files >1MB. Requires Python 3.10+.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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
- Software
Python
- Optional
Git
Cross-platform (Windows, macOS, Linux). Zero third-party packages — runs offline on Python 3.10+ stdlib with zero network calls. Git CLI is only needed for staged and history scans.
Permissions
File Scopes
Read-only scanning of source and config files to detect leaked secrets; writes only when explicitly saving an attestation file (--attest-out). Zero network access or external telemetry.