The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze arbitrary files from a codebase, which represents an attack surface for indirect prompt injection. Maliciously crafted files in the repository could contain instructions that attempt to override the agent's documentation task or behavior.\n
Ingestion points: The agent is instructed to use Read and Glob tools to analyze source code, configuration files, and existing READMEs (SKILL.md).\n
Boundary markers: The instructions lack explicit directives or delimiters to ensure the agent ignores instructions potentially embedded in the files it analyzes.\n
Capability inventory: The skill possesses the ability to Write files to the local directory and execute a scoped set of Bash commands for project discovery (SKILL.md).\n
Sanitization: There are no instructions for sanitizing or validating data read from the codebase before it is processed.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute specific shell commands (tree, find, wc, cloc) to perform project discovery and metric calculation. These commands are restricted in the skill's metadata to limit the execution scope (SKILL.md).

smart-docs