onboard
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill performs automated analysis of arbitrary codebases, which involves processing untrusted external data that could contain malicious instructions.
- Ingestion points: The skill reads source files, dependency manifests (e.g., package.json, Gemfile), and configuration files (e.g., .env.example) using tools like read_file, grep, and glob as defined in SKILL.md and sub-agent playbooks.
- Boundary markers: There are no explicit instructions or delimiters provided to the sub-agents to isolate potentially malicious prompts embedded in the codebase from the documentation task.
- Capability inventory: The sub-agents can read project files and write documentation to the .claude/onboarding/ directory. The coordinator has the capability to spawn and manage multiple sub-agents.
- Sanitization: No explicit sanitization or filtering of codebase content is mentioned before it is processed or included in the generated documents.
- [COMMAND_EXECUTION]: The skill uses local filesystem exploration tools to understand the project structure.
- Evidence: The sub-agent playbooks (e.g., structure-guide.md) recommend using shell commands such as 'find' to map the directory tree and 'grep' to identify specific code patterns like React Server Components usage.
Audit Metadata