The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides structured protocols for using standard diagnostic tools such as ps, ls, cat, and grep to verify service availability and configuration. These commands are used solely to generate evidence for task completion.
[EXTERNAL_DOWNLOADS]: Includes procedures for validating the installation of development tools through official package managers like npm and pip. These actions are restricted to ensuring the presence of required components in the development environment.
[REMOTE_CODE_EXECUTION]: Instructions include the use of curl for testing API endpoints and performing service health checks. These operations are used strictly to verify the functional correctness of the code being implemented.
[SAFE]: The skill emphasizes 'Anti-Cargo-Cult' reasoning and 'Independent Verification' to prevent the introduction of unvetted code and to mitigate automation bias during high-stakes changes. It enforces a strict 'Iron Law' requiring failing tests before any code implementation.

proof-of-work