repo-context-packaging

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: Employs standard Unix utilities such as git, grep, sed, and wc to analyze the repository structure and filter out non-essential files.- [EXTERNAL_DOWNLOADS]: Recommends the use of third-party utilities like repomix (via npx) and files-to-prompt to automate the bundling process. These are well-known tools in the developer ecosystem.- [INDIRECT_PROMPT_INJECTION]: The skill defines a process that ingests untrusted data from repository files into an agent's context, creating a potential surface for indirect injection.
  • Ingestion points: Reads all tracked files in the repository using git ls-files (SKILL.md).
  • Boundary markers: Uses Markdown headers and code blocks as delimiters in the output bundle.
  • Capability inventory: The skill uses shell execution and local file access (SKILL.md).
  • Sanitization: No specific instructions are provided to sanitize or escape malicious prompts embedded in the repository's source code or documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 06:31 PM
Security Audit — agent-trust-hub — repo-context-packaging