learn-codebase
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell utilities such as
rg,grep, andgitto perform searches for project documentation and security anti-patterns within the codebase. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes instructions from untrusted external files (e.g.,
CLAUDE.md,.cursorrules) that comprise the codebase being analyzed. - Ingestion points: Steps 2 and 5 read content from project-specific instruction files and any file matched by the security scan.
- Boundary markers: The skill does not define specific delimiters or instructions to prevent the agent from obeying instructions embedded within the files it reads.
- Capability inventory: The agent has the ability to execute shell commands and modify the
.pi/settings.jsonfile. - Sanitization: No sanitization or escaping of the ingested file content is performed before it is presented to the user or processed by the agent.
Audit Metadata