The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the web during its research and validation phases, which could contain malicious instructions.
Ingestion points: The skill retrieves external data via WebFetch and WebSearch, and processes output from sibling tools like researchkit and validatekit in SKILL.md (e.g., under the research and validate modes).
Boundary markers: The skill instructs the agent to draft content and ask for user confirmation before writing, but it lacks specific instructions to delimit or sanitize external content to prevent the agent from obeying embedded instructions.
Capability inventory: The agent has powerful capabilities including Bash, Write, Edit, and WebFetch across all modules.
Sanitization: No explicit sanitization or filtering of external inputs is mentioned.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to manage the local environment, specifically for initializing the ideas repository with git init as described in the "The ideas repo" section of SKILL.md.
[EXTERNAL_DOWNLOADS]: The skill performs network operations via WebFetch and WebSearch to fulfill research and validation requests, which is a core feature of the tool's intended purpose to gather external context for user ideas.

ideakit