The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code changes from git diffs, which presents a potential surface for indirect prompt injection where instructions could be hidden in comments or code, though this is inherent to the skill's primary purpose as a code simplification utility.
Ingestion points: Reads external data via git diff output to identify code units for ablation (SKILL.md).
Boundary markers: None present to isolate diff content from the agent's internal instruction context.
Capability inventory: Executes git commands and a project-local check script for verification (SKILL.md).
Sanitization: No sanitization or escaping of the diff content is performed before processing.

simplify