Dependency License Checker Skill

    1

    Find the dependencies with risky licenses before your customers, auditors or acquirers do. Works offline, from data already on your disk.

    Free

    2 uses5.0 (1 review)

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Dependency License Checker Skill

    Example session with this skill installed

    Audit my node_modules folder for a distributed proprietary app. Generate a report and a notices file, then tell me which packages are 'Blocked'.

    • Read your context and instructions
    • Compiled the dependency license checker

    I have completed the license audit of your node_modules directory. I assumed a

    distributed software model for a

    proprietary project. The scan processed 142 dependencies.

    Summary of Findings

    • Blocked: 2
    • Avoid: 1
    • Review: 5
    • OK: 134

    License Audit Report

    PackageVersionLicenseVerdictReason
    fast-glob-sync1.2.0GPL-3.0-onlyBlockedCopyleft license in distributed proprietary app.
    internal-ui-lib0.4.2UNKNOWNBlockedNo license file or metadata found; default is no permission.
    heavy-auth-helper2.1.0AGPL-3.0-or-laterAvoidStrong network-copyleft; requires legal review for distribution.
    utils-lite1.0.1BSD-3-ClauseOKPermissive license compatible with your model.

    I have generated THIRD_PARTY_NOTICES.md containing the required attribution text for the 134 'OK' and 'Review' components. Note that internal-ui-lib was omitted from the notices as no license text was found on disk.

    Disclaimer: This report is for triage purposes only and is not legal advice.

    Next steps
    1.

    Replace fast-glob-sync: This is a GPL-licensed component in a distributed proprietary app. I can suggest MIT/Apache-2.0 alternatives.
    2.

    Identify internal-ui-lib: Check the repository for this package to find a LICENSE file and manually update the policy.
    3.

    Legal Review: Send the 'Avoid' finding for heavy-auth-helper to your legal contact to determine if your linking method triggers copyleft requirements.

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

    About this skill

    One GPL library in a closed-source product, one AGPL service dependency, one package with no license at all: these are the findings that surface during due diligence, enterprise security reviews and acquisitions. This skill helps your agent find them first.

    A bundled scanner reads license data that already exists on your machine: installed packages (a node_modules folder or a Python virtualenv), package-lock.json, CycloneDX and SPDX SBOMs, and exports from license-checker, pip-licenses, cargo-license and go-licenses, or any CSV with name, version and license. Nothing is looked up online.

    Each license is normalized to SPDX, placed in a risk category, and given a verdict (ok, review, avoid, blocked) for how you actually ship: hosted service, shipped software, published library, internal tool, or open source (with GPL, AGPL, LGPL, MPL and Apache compatibility checks against your own license). It resolves OR and AND expressions and linking exceptions, recovers a missing license from the license file's text, flags a license file that looks stricter than the declared license, and treats a missing license as blocked instead of assuming it is fine.

    Turn findings into action: an allow list, deny list and recorded exceptions (with approver and expiry) in a JSON policy file that works the same in CI, a --fail-on exit code, and a generated third-party notices file built from the license files it found.

    The agent is told never to state a package's license from memory, to ask how the software is shipped before judging, and to treat every result as triage, not legal advice. Includes guides to license categories and obligations, getting license data for each ecosystem, and acting on findings (options, review questions, ticket and maintainer-message templates).

    Good to know: Python 3.9+ and nothing else. No network access, no environment variables read, and your lock files and packages are never modified. Scanned 6,000 packages in about a second in testing. Not legal advice; have your legal contact review anything flagged avoid or blocked.

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

    Requires local package data or SBOMs. Does not scan binary files, images, or vendored code. Logic covers common OSS licenses only. Triage only; not legal advice.

    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

    5.0
    1 review
    5
    1
    4
    0
    3
    0
    2
    0
    1
    0

    This skill has been used 2 times.

    Trust & safety

    Security scan

    Security scan passed 23 days ago

    • Free with an Agensi account

    Permissions

    Read Files
    Write Files
    Terminal / Shell
    Listed23 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard LLMs that might guess licenses based on training data, this skill uses a local Python scanner to extract data directly from your project files and lockfiles, ensuring your compliance report matches the exact version of the code you are shipping.

    The skill is compatible with Node.js, Python, Go, and Rust, and can also process standard CycloneDX or SPDX SBOM formats to audit projects in other languages.

    Yes, you can define your distribution model (such as SaaS or distributed library) and set custom allow/deny lists to automatically flag specific licenses like GPL or AGPL based on your company's risk tolerance.

    The package includes the core Python-based scanner, pre-configured prompts for risk triage and remediation, and templates for generating standardized THIRD_PARTY_NOTICES files.

    The skill is designed for any AI agent with local file system access, as it requires the ability to execute the Python 3.9+ scanner against your repository's manifests and metadata.

    New in Legal & Compliance

    Free