cortex-m-debug

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, exfiltration attempts, or unauthorized behaviors were detected. The skill's operations are consistent with its stated purpose of assisting with embedded firmware debugging.
  • [COMMAND_EXECUTION]: The skill provides standard command-line examples for common, well-known debugging tools such as GDB, OpenOCD, pyOCD, and J-Link. These are intended for local use by the user in a development environment.
  • [PROMPT_INJECTION]: The skill includes a 'Safety Rules' section that correctly instructs the agent to avoid performing destructive actions (like mass erasing or changing security configurations) without explicit user approval. This demonstrates good safety alignment.
  • [DATA_EXFILTRATION]: No network exfiltration or unauthorized sensitive file access was found. The skill operates on technical artifacts like ELF files, linker scripts, and memory registers relative to the target microcontroller.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 07:01 AM
Security Audit — agent-trust-hub — cortex-m-debug