agent-protocol

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill defines a logical framework and syntax for agent interaction without using executable code or system-level commands. It introduces structural constraints, such as maximum call depth and circular call prevention, which contribute to the stability and predictability of agent interactions.
  • [NO_CODE]: All files analyzed are documentation and instruction-based markdown files; no scripts (e.g., Python, JavaScript, Shell) or binaries are present.
  • [PROMPT_INJECTION]: The protocol uses structured tags like [INVOKE] and [RESPONSE] to facilitate communication but does not contain directives aimed at bypassing safety guardrails, ignoring instructions, or extracting system prompts.
  • [DATA_EXFILTRATION]: The skill does not contain any hardcoded credentials, sensitive file path references, or network-bound operations.
  • [COMMAND_EXECUTION]: No shell commands or dynamic context injection patterns (such as !command) were identified in the provided files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 03:02 AM
Security Audit — agent-trust-hub — agent-protocol