The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill directs users to install software from an unverified third-party repository (github.com/warmskull/idea-darwin) and a non-standard package registry (ClawHub). These sources are not recognized as trusted vendors or well-known services.
[COMMAND_EXECUTION]: The documentation references the use of custom slash commands (e.g., /idea-darwin init, /idea-darwin round) which imply the execution of underlying scripts or binaries to perform 'evolution' tasks.
[PROMPT_INJECTION]: The skill exhibits surface area for indirect prompt injection by ingesting and processing untrusted data from user-created files (ideas.md and stimuli.md).
Ingestion points: Reads content from ideas.md and stimuli.md to perform 'structured research' and 'deepening'.
Boundary markers: None identified in the provided instructions to separate user data from agent instructions.
Capability inventory: The agent uses the content to generate briefings and make 'life-or-death' decisions on idea evolution via CLI commands.
Sanitization: No evidence of input validation or instruction filtering for the processed markdown files.

idea-darwin