The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill utilizes the Bash tool to run existing test suites during the 'Establish Baseline' and 'Execute Refactorings' phases. This is a standard and necessary operation for verifying behavior preservation in a refactoring workflow and aligns with the skill's primary purpose.\n- [INDIRECT_PROMPT_INJECTION] (LOW): The skill is designed to ingest and analyze external source code, which serves as a potential vector for indirect prompt injection if an attacker embeds malicious instructions within the codebase being refactored.\n
Ingestion points: Local source files accessed via Read, Grep, and Glob tools during analysis and planning phases.\n
Boundary markers: None identified; the agent processes code content as data to identify structural 'smells'.\n
Capability inventory: Includes Bash for command execution, Edit and TodoWrite for file modification, and Read for file access.\n
Sanitization: No explicit sanitization or filtering of instructions is performed on the ingested code before processing.

safe-refactoring