Neural Network Ops
v1.0.0Diagnoses and tunes LLM providers (Groq, OpenRouter, Ollama), resolves rate limits/timeouts, and selects stable primary/fallback models. Use when the bot is...
⭐ 0· 142·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description claim provider/gateway diagnostics and routing; the SKILL.md contains exactly the expected checks (journalctl, systemctl), routing priorities, fallback config examples, and recovery steps for OpenClaw, Groq, OpenRouter, and Ollama — these are proportionate to the stated purpose.
Instruction Scope
Instructions are specific and focused on service health and model routing. However they include destructive/privileged actions (rm -rf of session files under /root/.openclaw, systemctl restart/stop/start) and direct curl calls to a local API. These are in-scope for operations but require operator confirmation and appropriate privileges; the instructions assume root/systemd and a particular filesystem layout.
Install Mechanism
No install spec or code files are present (instruction-only). This minimizes supply-chain risk; nothing is downloaded or written by the installer.
Credentials
The skill declares no required environment variables, credentials, or config paths. The SKILL.md also does not attempt to read external secrets or request API keys — consistent with its purpose of local ops and routing configuration.
Persistence & Privilege
The skill does not request permanent presence (always:false) and is user-invocable. Nevertheless, its recommended actions require elevated privileges (systemctl, modifying /root files). Ensure the agent is not given blanket root execution rights or allowed to run these commands without explicit human confirmation.
Assessment
This instruction-only skill appears to be what it says: an ops playbook for OpenClaw model providers. It does include commands that restart services and delete session files under /root, so only run its steps when you understand the consequences and have backups. Recommended precautions: 1) Use it in staging first; verify paths and service names match your installation. 2) Require explicit human confirmation before executing restart or rm commands. 3) Do not grant the agent automatic root/sudo rights; run commands under an account with least privilege or via an approved runbook. 4) Back up session/config files before applying deletions or replacements. 5) If you need the agent to act automatically, add granular guards (whitelists, prompt confirmations) — otherwise keep it user-invocable only.Like a lobster shell, security has layers — review code before you run it.
latestvk9741fs07mwhsfwn4e8caqpztn83dzfc
License
MIT-0
Free to use, modify, and redistribute. No attribution required.