The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes responses from the Higress Console API (e.g., consumer lists, AI routes, and providers) which represents an ingestion surface for untrusted data if the gateway settings are misconfigured.
Ingestion points: API responses fetched via curl from endpoints such as http://127.0.0.1:8001/v1/consumers and http://127.0.0.1:8001/v1/ai/routes (SKILL.md).
Boundary markers: Absent. The agent is not instructed to use delimiters or treat the API output as untrusted data.
Capability inventory: The skill uses curl for network interaction, jq for data manipulation, and openssl for credential generation (SKILL.md).
Sanitization: Absent. The skill uses jq for structured data handling but does not perform explicit sanitization of string content before processing.
[COMMAND_EXECUTION]: The skill relies on shell commands to perform gateway management tasks, which is consistent with its primary purpose.
Evidence: Numerous bash examples in SKILL.md demonstrate using curl for API interactions, jq for JSON manipulation, and openssl rand for generating security keys.

higress-gateway-management