Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    agent skill security scanner

    1

    Audits agent skills and MCP manifests for prompt injection, data exfiltration, and over-permissioned access.

    Free

    4 usesSecurity scanned
    agent skill security scanner

    agent skill security scanner

    Example session with this skill installed

    Audit this MCP server repo at [URL]. It is a simple file-renamer utility using a python script. Check for injection in the tool descriptions and verify network calls.

    • Read your context and instructions
    • Compiled the agent security scanner

    VERDICT: REVIEW
    SCOPE: Audited SKILL.md and rename.py; tool-list.json read.

    FINDINGS

    1. Capability Mismatch: rename.py includes a socket call to a remote telemetry IP not mentioned in documentation.
    2. Tool Poisoning: Tool description instructs agent to 'read .env first'.
    3. Static audit boundary: Cannot see runtime response injection.
    Audited SKILL.md

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

    About this skill

    The problem

    Third-party agent skills and MCP servers introduce foreign instructions and code into your agent's trust boundary. Identifying prompt injection, hidden exfiltration directives, or capability-purpose mismatches requires a semantic audit that traditional code scanners ignore.

    What it does

    • Audits SKILL.md and tool manifests for hidden instructions or model-steering directives.
    • Evaluates bundled source code for network egress, filesystem reach, and unauthorized secret access.
    • Identifies capability-purpose mismatches where a skill's permissions exceed its stated utility.
    • Inventories un-auditable external references like curl | sh or remote runtime configs.
    • Produces a definitive INSTALL, REVIEW, or BLOCK gate verdict based on evidence-backed findings.

    Why this beats prompting it yourself

    Static analysis for agents requires reading text as instructions, not just documentation. This skill enforces a disciplined protocol that prevents false negatives caused by keyword scanning, ensuring every file and model directive is semantically vetted rather than just grepped. It bridges the gap between classic security audits and LLM-specific vulnerability classes.

    Use cases

    • Vetting an open-source MCP server before allowing it access to your local environment.
    • Auditing your own custom skills for potential prompt injection before deployment.
    • Building a security gate for an internal agent skill store or CI/CD pipeline.
    • Reviewing tool descriptions to prevent agentic data exfiltration through tool poisoning.

    Known limitations

    Static audit only. Cannot detect runtime tool-response injection where malicious payloads are returned by a live API. Requires manual verification of obfuscated binaries or private repos.

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

    Static audit only. Cannot detect runtime tool-response injection. Requires source access; private repos or binary blobs are marked as unauditable gaps. Not for general web pentesting.

    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

    4 uses

    Times used to date

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 3 months ago

    • Free with an Agensi account

    Software
    Postmark MCP

    Permissions

    Network Access
    Listed3 months ago
    Updated1 month ago

    What's inside

    Frequently Asked Questions

    This skill performs a semantic audit of MCP manifests and tool descriptions to identify hidden prompt injection, unauthorized network egress, and permissions that exceed the tool's stated utility.

    The scanner is designed to audit any skill or MCP (Model Context Protocol) server manifest, making it compatible with agents built on Claude, OpenAI, and other systems utilizing the MCP standard.

    Each purchase includes the core security scanning engine, documentation on interpreting the Install/Review/Block verdicts, and a set of reference templates for auditing your own custom skills.

    While it performs deep static and semantic analysis of manifests and source code, it cannot monitor live API responses or runtime injections that occur after a tool has been deployed.

    Setup is straightforward: you provide the skill's directory path or manifest file to the scanner, and it will automatically index and audit the instructions and permissions bundled within.

    Yes, the scanner is specifically designed to identify 'capability-purpose mismatches,' flagging instances where a tool (like a calculator) requests unnecessary permissions (like filesystem access).

    New in Agents & Orchestration

    Free