The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documents the installation of the Supabase CLI through legitimate package managers, specifically npm (npm install supabase) and Homebrew. These are well-known developer services, and the referenced resources are standard dependencies for the skill's database functionality.
[COMMAND_EXECUTION]: The instructions direct the agent to interact with the fit-map CLI and the supabase stack. These commands—such as validate, generate-index, migrate, and activity start—are used to process local configuration files and manage a local development database, which is consistent with the primary purpose of defining engineering standards.
[PROMPT_INJECTION]: The skill processes external data that could theoretically harbor indirect instructions. However, the skill incorporates validation mechanisms and follows the vendor's structured schema, which minimizes the risk of unintended behavior.
Ingestion points: Local YAML definitions in data/pathway/, user-provided people files (YAML/CSV), and API payloads fetched from GetDX snapshots.
Boundary markers: The skill employs structured YAML with distinct sections (human: and agent:) and requires validation against JSON Schemas, providing structural boundaries for ingested data.
Capability inventory: The agent can execute local CLI commands via npx and perform file-system operations limited to the skill's data and configuration directories.
Sanitization: Ingested content is subjected to JSON Schema validation and referential integrity checks using tools from @forwardimpact/libskill.
[SAFE]: The skill adheres to security best practices by instructing users to manage secrets, such as GETDX_API_TOKEN, via environment variables rather than hardcoding them within the project files.

fit-map