The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Node.js script (generate-tech-specs.mjs) for automating the generation of technical documentation. The script includes a security function assertInsideRepo that validates paths, effectively preventing path traversal attacks by ensuring operations stay within the repository boundary.
[DATA_EXPOSURE]: The generator contract explicitly forbids the inclusion of secrets. Step 4 of the verification process requires the agent to confirm that no connection strings, credentials, tokens, or customer data are encountered or emitted during the derivation process.
[PROMPT_INJECTION]: The skill uses instructional framing to ensure execution integrity (e.g., 'Execute skill steps in declared order', 'NEVER skip, reorder, or merge steps'). These instructions are intended to maintain the 'Single Writer' and 'Idempotency' contracts of the documentation system rather than to bypass safety guardrails.
[SAFE]: The skill does not perform network operations, use obfuscation, or establish persistence. Its logic is focused on mechanical derivation (grepping for markers like TestSpec) and instantiating fixed markdown templates, which limits the risk of unpredictable behavior or data exfiltration.

tech-spec