The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies an indirect prompt injection attack surface by processing data from local codebase files that could contain malicious instructions.
Ingestion points: Untrusted file contents are read and returned to the agent context via grep and ast-grep searches.
Boundary markers: The skill lacks explicit instructions or delimiters to isolate search results from the agent's internal operational logic.
Capability inventory: The skill enables the execution of several shell commands, including ast-grep, grep, find, git, and bash.
Sanitization: While instructions provided include shell-level escaping of metavariables (e.g., using \$VAR), there is no mechanism to sanitize or filter the content ingested from the codebase.
[COMMAND_EXECUTION]: The skill relies on the execution of multiple system commands like ast-grep, grep, find, and git for its core functions. It provides instructions and examples for the agent to execute these tools directly.

godgrep