Caveat verdict
ollama-task-orchestrator
Remote Ollama task orchestrator via SSH to a user-configured worker machine; exec-arbitrary-command is a documented feature operating on the user own worker node.
⚠ 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.
Findings (7)
Accesses .ssh directory
README-claude-code.md · code · .ssh/
Pipe-to-python pattern — remote code execution risk
runner/run_task.sh · prose · downgraded · curl -s --max-time 3 "$api_base/api/ps" 2>/dev/null | python
Pipe to python — executes piped content as Python code
runner/run_task.sh · prose · downgraded · | python3
Accesses shell history/config
install.sh · prose · downgraded · ~/.zshrc
subprocess execution — runs system commands from Python
skill.py · prose · downgraded · subprocess.run(
References agent configuration files
README-claude-code.md · prose · downgraded · CLAUDE.md
Python os.environ.get — reads environment variable
skill.py · prose · downgraded · os.environ.get(
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B/C/D). 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.
network_in Is this flag fair?
Thanks — recorded.