bitflow
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external API endpoints (Bitflow BFF) and processes user-supplied JSON for subcommands such as add-liquidity-simple. While this creates a surface where external content could potentially contain instructions aimed at the agent, the skill utilizes structured JSON parsing and defined output mapping to mitigate the risk of untrusted data influencing the agent's behavior.
- [COMMAND_EXECUTION]: The skill executes write operations on the Stacks blockchain, including token swaps and liquidity provisioning. Security is maintained through local wallet management and integrated safety gates that require explicit user confirmation for trades with high price impact (exceeding 5%).
Audit Metadata