agents-panel
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains an indirect prompt injection vulnerability surface where untrusted data is directly interpolated into agent prompts.
- Ingestion points: User-supplied input via the
[problem or decision to analyze]argument and the{context}variable derived from conversation history are used in sub-agent prompts in the 'Mode A: Debate' (Section A3) and 'Mode B: Poll' (Section B2) sections. - Boundary markers: The prompts lack explicit boundary markers (such as XML tags or triple quotes) or specific 'ignore embedded instructions' warnings around the interpolated variables to differentiate user input from system-level instructions.
- Capability inventory: The skill has access to several powerful tools, including
Bash,WebSearch,WebFetch,Read,Grep, andGlob, as defined in the YAML frontmatter. - Sanitization: There is no implementation of escaping, validation, or filtering of the external content before it is processed by the sub-agents.
Audit Metadata