bk-cli-aidev
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform administrative tasks by executing shell commands using the bk-cli aidev tool. This includes operations for resource querying, creation, and management, as well as local file uploads.
- [INDIRECT_PROMPT_INJECTION]: The skill enables the agent to ingest data from the AIDev platform (such as knowledge base content, prompt collections, and agent configurations), creating a potential surface for indirect prompt injection. Ingestion points: Data is retrieved via query_knowledgebase, preview_knowledge, get_agent, and get_collection. Boundary markers: No explicit delimiters or instructions are provided to ignore embedded instructions. Capability inventory: The agent can execute shell commands through the bk-cli tool based on processed data. Sanitization: No instructions for sanitizing or validating platform content are present.
Audit Metadata