The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (source code) and instructs the agent to apply fixes, creating a vulnerability where instructions embedded in the code could influence agent behavior.
Ingestion points: The skill reads files identified by the git diff command in the SKILL.md file.
Boundary markers: There are no instructions provided to the agent to treat the code content strictly as data or to ignore embedded instructions.
Capability inventory: The skill requires the agent to "Apply fixes," which necessitates file system write capabilities.
Sanitization: No sanitization or validation of the source code content is performed before processing.
[COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands to discover files for analysis.
Evidence: Step 1 of the Process section in SKILL.md instructs the agent to "use git diff --name-only HEAD~1 or staged changes."

typo3-simplify