ClawAudit verdict
alphaear-predictor
Market time-series forecasting skill using a local Kronos model with explicit security caution about only placing trusted model weights in the models directory; all processing is local and the executionSinkDetected flag reflects documented Python script execution.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (5)
Uses eval() — can execute arbitrary code
scripts/utils/predictor/evaluation.py · prose · downgraded · eval(
Pipe to sh — executes piped content as shell commands
scripts/utils/stock_tools.py · prose · downgraded · |SH
Popular HTTP library — network access
scripts/predictor/model/kronos.py · prose · downgraded · got
Python os.getenv — reads environment variable
scripts/utils/llm/capability.py · prose · downgraded · os.getenv(
Python os.environ.get — reads environment variable
scripts/utils/stock_tools.py · prose · downgraded · os.environ.get(
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions — minimal attack surface.
Is this flag fair?
Thanks — recorded.