skills/time-attack/gstack/make-pdf/Gen Agent Trust Hub

make-pdf

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill downloads a manifest and corresponding runtime components from the vendor's official repository. It executes a setup script contained within these downloaded artifacts to configure the local environment. This process is secured via mandatory SHA-256 verification and optional Cosign attestation.- [EXTERNAL_DOWNLOADS]: Fetches binary artifacts and manifests from official GitHub release hosts. The download logic enforces strict host allow-listing and validates the final resolution URL before initiating transfers.- [DYNAMIC_EXECUTION]: Uses a bootstrap utility to dynamically locate and execute Node.js installer scripts from the local filesystem after a verified download has completed.- [INDIRECT_PROMPT_INJECTION]: The skill processes external Markdown files, which could theoretically contain malicious instructions targeting the agent context.
  • Ingestion points: Markdown source files processed by the generate and preview tools.
  • Boundary markers: The skill does not implement specific delimiters to isolate user-provided Markdown content from agent instructions during processing.
  • Capability inventory: Includes shell command execution, filesystem writes, and optional network fetching.
  • Sanitization: Content is rendered through standard engines like Paged.js and Mermaid, which provide structural constraints.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 11:22 PM
Security Audit — agent-trust-hub — make-pdf