NYC

pw-embedded-c-style

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [CREDENTIALS_UNSAFE] (SAFE): No hardcoded secrets, API keys, or private keys were detected. The file focuses purely on microcontroller register definitions (e.g., P0, TMOD).
  • [EXTERNAL_DOWNLOADS] (SAFE): The skill does not reference external package managers (npm, pip) or attempt to download scripts from the internet.
  • [REMOTE_CODE_EXECUTION] (SAFE): There are no commands for remote script execution or local subprocess spawning. The C code provided is intended for target hardware, not for execution on the host system.
  • [DATA_EXFILTRATION] (SAFE): No network-capable commands (curl, wget) or sensitive file path access (e.g., ~/.ssh) were found.
  • [PROMPT_INJECTION] (SAFE): The instructions are technical and descriptive. There are no attempts to bypass safety filters or ignore prior instructions.
  • [INDIRECT_PROMPT_INJECTION] (SAFE): While the skill ingests code for optimization, it lacks any dangerous capabilities (file writes, network, or subprocesses) that could be exploited by malicious code embedded in the input.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:46 PM