The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the repository, including code diffs and local configuration files like AGENTS.md or repository-specific simplify skills.
Ingestion points: The "Core Loop" and "Read This First" sections in SKILL.md specify that the agent should read changed files, diffs, and local documentation to inform its actions.
Boundary markers: The instructions do not define specific delimiters or guardrails to ensure the agent ignores malicious instructions that might be embedded in comments or strings within the files it is simplifying.
Capability inventory: The skill has the ability to read files, execute local terminal commands for verification (lint/typecheck), and launch subagents using the Task tool to distribute work.
Sanitization: No sanitization or filtering logic is present to handle potentially malicious content in the files being processed.
[COMMAND_EXECUTION]: The skill explicitly instructs the agent to perform local command execution to verify that its simplifications did not break the code.
Evidence: The "Verification" section in SKILL.md directs the agent to "run the smallest relevant verification," such as typecheck or lint commands. While these are standard developer tools, they represent an execution capability within the local environment.

simplify