The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation recommends installing the @apify/mcpc package. This is the official Model Context Protocol server for Apify, a well-known and trusted service provider, ensuring the dependency is legitimate.
[COMMAND_EXECUTION]: The workflow utilizes the mcpc CLI tool and a provided Node.js script (run_actor.js) to interact with Apify Actors. These operations are scoped to the skill's primary function of data retrieval and analysis.
[CREDENTIALS_UNSAFE]: The skill requires an APIFY_TOKEN stored in a .env file. The implementation uses Node.js's native environment file support to load this secret securely, avoiding hardcoded credentials in the source code.
[PROMPT_INJECTION]: The skill processes social media data fetched from external sources through the Apify platform.
Ingestion points: Data items are retrieved from the Apify dataset API (api.apify.com) in reference/scripts/run_actor.js.
Boundary markers: The output does not use explicit boundary markers for the data presented to the agent, but it is structured by the script's logging functions.
Capability inventory: The run_actor.js script has file-writing capabilities (writeFileSync) and the agent is instructed to execute local scripts and CLI tools.
Sanitization: The script performs sanitization by truncating long text fields to 100-200 characters and applying standard CSV escaping to all output values, reducing the risk of downstream instruction injection or formatting exploits.

apify-influencer-discovery