The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary documentation (SKILL.md) and several provider-specific guides (e.g., LinkedIn) provide numerous Python code snippets using shell heredocs. These are designed to be executed by the agent to perform authentication, manage connections, and call native API endpoints through the vendor's gateway.
[PROMPT_INJECTION]: The skill facilitates the ingestion of data from a vast array of external APIs, which inherently introduces a risk of indirect prompt injection. Adversarial content within API responses could attempt to manipulate the agent's logic or bypass its safety protocols.
Ingestion points: Responses from any of the 100+ supported third-party APIs retrieved via the gateway.maton.ai proxy, as documented in SKILL.md and the references/ subdirectory.
Boundary markers: Absent. The provided instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes data returned from external APIs.
Capability inventory: The agent is instructed to use Python (urllib.request) and shell commands to interact with network resources and process external data structures.
Sanitization: Absent. The skill does not provide or suggest methods for sanitizing, escaping, or validating external content before it is interpolated into the agent's context.

api-gateway