The Agent Skills Directory

[CREDENTIALS_UNSAFE] (HIGH): The skill accesses the .env file, which is a sensitive file path, and uses grep to extract the APIFY_TOKEN. This practice exposes credentials to the shell environment and potential process logging.
[COMMAND_EXECUTION] (HIGH): Step 2 and Step 4 construct shell commands that interpolate environment variables (APIFY_TOKEN) and user-provided inputs (ACTOR_ID, JSON_INPUT). This structure is highly vulnerable to command injection if inputs are not strictly validated.
[EXTERNAL_DOWNLOADS] (MEDIUM): The workflow requires the global installation of the @apify/mcpc package. Since 'Apify' is not listed as a trusted organization in the security scope, this dependency is treated as unverifiable.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill dynamically fetches and processes Actor schemas and README content from the external Apify registry.
Ingestion points: Data is retrieved from mcp.apify.com using the mcpc tool in Step 2.
Boundary markers: Absent; the external content is piped directly to jq and presented to the agent without delimiters.
Capability inventory: The skill has capabilities for shell command execution, file system access (reading .env), and running local Node.js scripts.
Sanitization: No sanitization or validation is performed on the retrieved actor details before they are processed by the agent.

apify-content-analytics