The Agent Skills Directory

[SAFE]: No security issues detected. The skill uses standard utilities like find, grep, jq, and git to map codebase topology, dependencies, and conventions as intended for code analysis.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads and processes external source code and configuration files. Ingestion points: The agent reads files across the project directory including source code and build configurations (SKILL.md). Boundary markers: No explicit delimiters or ignore instructions are provided to separate untrusted file content from the agent's instructions. Capability inventory: Capabilities are limited to read-only shell operations (find, ls, cat, grep, jq, sed, git, wc); no network or file-write permissions are used. Sanitization: Information is partially sanitized via grep and sed filtering before the agent processes it.

code-research