The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to automatically perform an environment scan upon activation to check for recommended companion skills.
Evidence: The instruction in SKILL.md directs the agent to run ls ~/.claude/skills/ ~/.agent/skills/ ~/.agents/skills/ .claude/skills/ .agent/skills/ .agents/skills/ 2>/dev/null to identify installed skills.
[PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection as it is designed to process and review user-provided technical documentation, API specifications, and runbooks without explicit boundary markers or sanitization logic.
Ingestion points: User-provided documentation, architecture designs, and API specs processed in SKILL.md and the references/ directory.
Boundary markers: Absent. The skill does not define specific delimiters (e.g., XML tags) for untrusted data.
Capability inventory: The agent retains the ability to execute shell commands and read filesystem paths as demonstrated by the companion check instructions.
Sanitization: Absent. No instructions are provided to the agent to escape or validate external content before processing.
[EXTERNAL_DOWNLOADS]: The skill directs the user and the agent to install additional components using the npx skills utility from the author's repository.
Evidence: Instructions in README.md and SKILL.md recommend running npx skills add AbsolutelySkilled/AbsolutelySkilled --skill <name> to extend capabilities.

technical-writing