peas-example-coach
Pass
Audited by Gen Agent Trust Hub on Apr 15, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill reads from a local file
example-learning-checklist.mdand writes records to thesession-records/directory. These operations are restricted to the local workspace and are necessary for the skill's primary function as a learning coach. No sensitive system files or credentials are targeted. - [COMMAND_EXECUTION]: While a Python code snippet is provided in
references/peas-splash.md, it is included as a static reference for developers to implement terminal-based splash screens with ANSI colors. The agent is not instructed to execute this code or any other external scripts dynamically. - [PROMPT_INJECTION]: The skill contains extensive behavioral instructions (e.g., 'Internal preparation must not enter the screen', 'Forbidden to use internal monologue wrappers'). These are design constraints for the 'Coach' persona to maintain a clean user interface for students, rather than attempts to bypass host security filters or extract system prompts.
- [SAFE]: No network operations (
curl,wget,fetch) are present. All resource references target local markdown files within the skill's own directory structure or the user's project root.
Audit Metadata