The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of external data from a Qdrant database and inserts it into the prompt of a target agent, which is a standard pattern that introduces an indirect prompt injection surface.
Ingestion points: Data is ingested through the qdrant-find tool and stored in the content variable within the Enriched Prompt Template in SKILL.md.
Boundary markers: The skill uses markdown headers (e.g., ## From Qdrant) and horizontal separators (---) to delimit the retrieved context from the task, but does not provide explicit instructions to the downstream agent to treat the retrieved content as untrusted data.
Capability inventory: The skill utilizes data retrieval (qdrant-find), data storage (qdrant-store), and agent delegation (Task tool) in SKILL.md.
Sanitization: No sanitization, filtering, or validation of the content retrieved from the database is performed before it is presented to the target agent.

rag-wrapper