bd-go
Pass
Audited by Gen Agent Trust Hub on Apr 1, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to fetch tasks using the
bd readycommand and implement them. Since these tasks originate from an external source (a task tracker or project management tool), they represent an untrusted data entry point. An attacker could embed malicious instructions in a task description to influence the agent's behavior during implementation. - Ingestion points: Output of
bd ready(referenced in SKILL.md). - Boundary markers: Absent; there are no instructions to the agent to ignore or delimit instructions found within the tasks themselves.
- Capability inventory: The instruction to "implement it" suggests the agent will use its full suite of tools (file writes, shell execution) to fulfill the task.
- Sanitization: Absent; the skill does not suggest any validation or filtering of the task content.
- [COMMAND_EXECUTION]: The workflow involves claiming and implementing tasks, which inherently requires the agent to execute commands and modify the local environment. This is the intended purpose of the skill, but it serves as the primary impact vector if an indirect prompt injection occurs.
Audit Metadata