The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from the local codebase and git diffs, which are then passed to sub-agents for analysis and subsequent code modification.
Ingestion points: The skill captures data via git diff and allows sub-agents to use read_file and search_files to ingest the wider codebase.
Capability inventory: The agent has the ability to execute shell commands (git, tests, linters) and modify the filesystem via patch and write_file tools.
Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore potential embedded commands when passing the diff content to sub-agents.
Sanitization: There is no evidence of sanitization or filtering of the code content before it is processed by the LLM sub-agents.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands to function.
Evidence: It uses git diff variants to identify changes and instructs the agent to run project-specific tests and linters to verify the applied fixes. While these are standard developer tools, they represent an interaction with the host environment.

simplify-code