The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute standard git commands (git status, git commit). These operations are used for repository maintenance before and after analysis and are gated by user confirmation via AskUserQuestion.
[DATA_EXFILTRATION]: No network activity or unauthorized data access was detected. The skill reads local project files and writes its findings to a report within the local .agents/research/ directory.
[PROMPT_INJECTION]: The skill scans external source code, creating a potential surface for indirect prompt injection. However, it contains no instructions to bypass the agent's core safety filters or manipulate its system prompt.
Ingestion points: Source code files are processed through Grep and Read operations in Step 2 and Step 3.
Boundary markers: None implemented for input data.
Capability inventory: Includes Bash, Write, and AskUserQuestion tools.
Sanitization: Content from scanned files is analyzed directly without prior filtering.

performance-check