skills/simota/agent-skills/Accord/Gen Agent Trust Hub

Accord

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for specification generation and cross-team alignment. It explicitly prohibits the creation of implementation code or visual artifacts, focusing entirely on documentation.
  • [DATA_EXPOSURE]: No hardcoded credentials, API keys, or sensitive system file paths were found. The skill maintains an activity log in the .agents/ directory, which is standard behavior for task-tracking and does not involve exfiltration.
  • [REMOTE_CODE_EXECUTION]: There are no commands or patterns associated with downloading or executing external scripts. It does not use curl, wget, or package managers like npm or pip.
  • [COMMAND_EXECUTION]: The instructions do not contain any system-level command execution, shell spawning, or privilege escalation attempts.
  • [PROMPT_INJECTION]: The prompt instructions are focused on workflow and output formatting. No jailbreak attempts, safety bypasses, or instructions to ignore previous system prompts were detected.
  • [NO_CODE]: The skill body and its reference files contain no executable scripts, strictly adhering to its stated purpose of not writing code.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 02:07 AM