The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes external log data, which presents a surface for indirect prompt injection where malicious instructions could be embedded in application logs to influence the agent.
Ingestion points: The skill frequently uses the list_logs tool (SKILL.md, references/log-analysis.md) to retrieve runtime and build output.
Boundary markers: The instructions do not specify strict delimiters or "ignore instructions" warnings when processing log content.
Capability inventory: The skill has access to sensitive operations including update_environment_variables, query_render_postgres, and the ability to modify local source code (SKILL.md).
Sanitization: No explicit sanitization or filtering logic is provided for the retrieved log data before analysis.
[EXTERNAL_DOWNLOADS]: The skill uses curl to fetch the latest documentation from Render's official site (e.g., https://render.com/docs/mcp-server.md). These references are used to maintain up-to-date platform knowledge and are directed at the vendor's own well-known documentation domain.

render-debug