The Agent Skills Directory

[DYNAMIC_EXECUTION]: Several JavaScript utility scripts including validate-with-schema.mjs, mobile-preview.mjs, and preview-with-playwright.mjs utilize dynamic module resolution. They programmatically identify potential installation paths and use createRequire to load dependencies such as ajv, qrcode, and playwright at runtime.\n- [COMMAND_EXECUTION]: The skill uses eval to parse and store output from its CLI tools as documented in media.md. It also relies on shell scripts (gates.sh, shoot.sh) that execute various commands to manage validation gates and automate browser-based screenshots.\n- [EXTERNAL_DOWNLOADS]: The skill environment setup involves downloading and installing Node.js packages from the NPM registry into local cache folders and fetching JSON schemas from adaptybuilder.com.\n- [INDIRECT_PROMPT_INJECTION]: The skill acts on flow configuration JSON files which are treated as untrusted data.\n
Ingestion points: Reads JSON configuration files from local paths (flow.working.json) or API responses.\n
Boundary markers: The skill does not explicitly use delimiters for all user data within the prompt instructions, although processing is delegated to specific parsing scripts.\n
Capability inventory: The skill can perform shell command execution, file system writes, and network requests via its associated scripts.\n
Sanitization: Implements internal validation logic via verify-config.py and validate-with-schema.mjs to check document integrity before and after edits.

flow-generator