The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data, such as support tickets and API responses, which introduces a surface for indirect prompt injection.\n
Ingestion points: The agent is instructed to parse support tickets pasted by users and ingest survey configuration JSON from the PostHog API (e.g., GET /api/projects/<id>/surveys/<sid>/).\n
Capability inventory: The agent has access to powerful PostHog MCP tools, including execute-sql, activity-log, and persons, which could be targeted by instructions hidden in external data.\n
Boundary markers: The instructions do not specify the use of clear delimiters or 'ignore' instructions when handling untrusted data from tickets or API payloads.\n
Sanitization: There are no explicit validation or escaping requirements defined for values extracted from external sources before they are interpolated into SQL templates or reasoning processes.\n- [EXTERNAL_DOWNLOADS]: The skill involves fetching configuration and event data from external PostHog API endpoints.\n
Evidence: References to official API paths such as GET /api/projects/<id>/surveys/ and dashboard URLs at https://<us|eu>.posthog.com/.\n
Note: These references target the vendor's own established infrastructure and well-known services, representing normal operational behavior for a diagnostic tool.

debugging-surveys