The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from database internal tables and job descriptions, which presents a potential surface for indirect prompt injection if those fields contain adversarial content.
Ingestion points: SKILL.md includes queries to crdb_internal.jobs, crdb_internal.ranges, and crdb_internal.gossip_nodes, where results are processed by the agent.
Boundary markers: No explicit delimiters or boundary markers are utilized to wrap retrieved data.
Capability inventory: The skill possesses extensive capabilities including SQL execution via cockroachdb-execute-sql and infrastructure management via the ccloud and cockroach CLI tools.
Sanitization: There is no evidence of specific sanitization or filtering applied to external data retrieved from the database.

cockroachdb-operator