yy-learn-project
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the codebase being analyzed, including source code, README files, and project-specific instructions (e.g., AGENTS.md). This creates a surface for indirect prompt injection, where malicious comments or documentation could influence the agent's logic or output. * Ingestion points: Project source code, manifest files (e.g., package.json, pyproject.toml), and documentation (AGENTS.md, README.md) as specified in Steps 2, 3, 4, and 5 of SKILL.md. * Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore control sequences within the read data. * Capability inventory: The skill performs file reads across the workspace, writes to the learning/ directory, and may execute project-provided validation commands. * Sanitization: Step 8 includes an explicit requirement for the agent to verify that generated documents do not leak secrets, environment variables, or sensitive configurations.
- [COMMAND_EXECUTION]: Step 8 of SKILL.md instructs the agent to execute Markdown validation commands provided by the target project. If a malicious repository contains a harmful script masked as a Markdown checker, it could lead to unauthorized command execution within the agent's environment.
Audit Metadata