Caveat verdict
audio-transcribe
zxkane-audio-transcriber-funasr
The skill transcribes audio files locally using FunASR/MiMo models with optional LLM cleanup via documented API keys (Anthropic, OpenAI, AWS Bedrock); credential access is clearly scoped and transparent, the executionSinkDetected flag likely refers to documented bash commands for running scripts, and there is no exfiltration or deceptive behavior.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Permission integrity
agent_memory
Findings (9)
Accesses AWS credentials file
SKILL.md · frontmatter · ~/.aws/credentials
Accesses cloud provider credentials
SKILL.md · frontmatter · ~/.aws
References sudo — requests elevated privileges
references/pipeline-details.md · code · sudo
Instructs covert action — may act without user awareness
references/pipeline-details.md · prose · downgraded · silently
subprocess execution — runs system commands from Python
scripts/mimo_asr.py · prose · downgraded · subprocess.run(
apt-get install — installs system packages
scripts/setup_env.sh · prose · downgraded · apt-get install
Python os.environ.get — reads environment variable
scripts/mimo_asr.py · prose · downgraded · os.environ.get(
Blob URL — may embed executable content
scripts/setup_mimo.sh · prose · downgraded · blob:
Popular HTTP library — network access
scripts/test_speaker_verification.py · prose · downgraded · got
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B/D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
Requires 2 system binaries.
credential_storeagent_memory Is this flag fair?
Thanks — recorded.