The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands such as pwd, ls, find, cat, and jq for project discovery and context gathering. These commands are limited to information retrieval and directory exploration within the local environment.
Evidence: Found in SKILL.md (First Steps section) and reference files (e.g., references/init.md, references/tasks.md).
[PROMPT_INJECTION]: The skill processes user input and existing project files (e.g., source code, requirements) to generate documentation, which represents an indirect prompt injection surface. The behavior is inherent to the skill's purpose and is managed through structured interactions.
Ingestion points: Reads project files including requirement.md, design.md, coding-rules.md, CLAUDE.md, AGENTS.md, and discovered source files.
Boundary markers: Absent.
Capability inventory: Shell commands for directory and file discovery (ls, find, cat, jq).
Sanitization: No specific content filtering or sanitization of ingested file content is implemented.

spec-generator