hermes-bot-team-design

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process an external "project specification" to generate bot roles, behaviors, and SOUL.md files. This creates a surface where adversarial instructions in the specification could influence the behavior of the resulting bots.
  • Ingestion points: The project specification document used in Step 1 of the design workflow (SKILL.md).
  • Boundary markers: The instructions lack explicit guidance on using delimiters or instruction-ignoring prompts when processing the external specification.
  • Capability inventory: The skill performs local directory creation (mkdir), file content generation (SOUL.md, config.yaml), and execution of local management tools (hermes CLI).
  • Sanitization: There are no specific steps provided to sanitize or filter the content of the project specification before it is used to define bot profiles.
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform operations on the local filesystem and interact with the hermes CLI tool.
  • Evidence: Creating directory structures using mkdir -p ~/workspace/MeMex-Zero-RAG/wiki/agentforge/group-chat/<project-name>/.
  • Evidence: Verifying profile existence and status using ls ~/.hermes/profiles/<name>/ and hermes gateway list.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 09:39 PM
Security Audit — agent-trust-hub — hermes-bot-team-design