The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as CI logs and Git commit metadata which may contain untrusted content.
Ingestion points: SQL queries in references/investigation-queries.md retrieve content from engineering_analytics_ci_failures (fields like error_signature and test_id) and engineering_analytics_ci_job_history (fields like commit_message).
Boundary markers: The skill lacks explicit instructions or markers to delimit external data from agent commands, which could allow malicious content in logs or commit messages to influence the agent.
Capability inventory: The skill employs the execute-sql tool to interact with the data warehouse.
Sanitization: No sanitization or validation logic is specified for the data retrieved before it is presented to the agent or user.

investigating-ci-failures