The Agent Skills Directory

[SAFE]: The skill implements a strictly linear workflow with "Hard Gates" to ensure that no code implementation or project scaffolding occurs until a design specification is formally reviewed and approved by the user.
[COMMAND_EXECUTION]: The skill uses standard shell commands like grep, glob, and ls to explore the project structure and check for the presence of recommended companion skills in local environment directories (e.g., ~/.claude/skills/).
[EXTERNAL_DOWNLOADS]: The documentation provides instructions for installing the skill and its dependencies from the AbsolutelySkilled repository using the npx skills command, which is a standard procedure for this platform.
[PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection as it ingests untrusted data from the local project (documentation and source code) to inform its brainstorming process.
Ingestion points: Phase 1 (Deep Context Scan) and Phase 2 (Codebase-First Intelligence) in SKILL.md involve reading the docs/ directory, README.md, and searching the codebase for patterns.
Boundary markers: Absent; the skill does not explicitly use delimiters or ignore instructions when processing the content of ingested files.
Capability inventory: The skill can write new design specification files to the docs/plans/ directory and prompt the user to trigger other automated workflow skills.
Sanitization: Absent; the skill synthesizes information from the codebase directly into its logic without explicit filtering of potential instructions.

super-brainstorm