The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is authorized to modify project-level instruction files (AGENTS.md, CLAUDE.md). These files are used by AI agents to define their operational rules and personas. This functionality is the primary purpose of the skill and is documented as a maintenance task.- [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by design, as it ingests user-supplied 'guidance' and persists it into the agent's core instruction set.
Ingestion points: User-provided writing preferences, rules, and constraints (documented in SKILL.md and guidance-rules.md).
Boundary markers: No specific delimiters are requested for the generated guidance content in the target files.
Capability inventory: File-write access to root guidance/instruction files.
Sanitization: The skill performs structural sanitization (compressing into short sentences, merging duplicates) but does not implement security-focused content filtering.

ns-guidance