The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes application logs and database query results, which are untrusted external inputs. These could contain malicious instructions designed to subvert the agent's behavior during debugging.
Ingestion points: The skill uses list_logs to retrieve application output and query_render_postgres to read database content in SKILL.md and references/database-debugging.md.
Boundary markers: No boundary markers or instructions to ignore embedded commands are present in the debugging workflows.
Capability inventory: The skill has the ability to write to the service configuration via update_environment_variables and execute data manipulation commands via query_render_postgres.
Sanitization: There is no evidence of sanitization or filtering of external content before it is analyzed or used to generate fixes.
[COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary SQL commands through the query_render_postgres tool to analyze database state. While intended for diagnostics, this tool enables high-privilege operations on the database.

render-debug