The Agent Skills Directory

[SAFE]: The skill utilizes well-known and trusted services including Anthropic, Firecrawl, and Supabase to perform its stated functions. These network operations are necessary for the quantity extraction and data logging features described.
[CREDENTIALS_UNSAFE]: The skill follows security best practices for secret management. It instructs users to store sensitive keys in environment variables and uses non-sensitive placeholders like 'YOUR_GITHUB_TOKEN' in documentation examples.
[COMMAND_EXECUTION]: Documentation for the skill includes a standard curl example for interacting with the GitHub Actions API. This is a routine automation practice and does not involve unsafe command injection or unauthorized execution.
[DATA_EXFILTRATION]: The skill logs construction project metadata and calculated costs to a Supabase database. This behavior is a core documented feature for tracking project history and does not constitute unauthorized data exfiltration.

construction-takeoff