Openclaw Llm Router
v2.1.0Intelligent model router for OpenClaw. Reads your local config, classifies available models into three tiers (Premium / Balanced / Economy), and routes each...
⭐ 0· 155·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill declares only python3 and provides a script that reads OpenClaw config files and environment keys to determine active providers and available models, which aligns with the router description. The providers and model-tier mappings in the references and script match the stated functionality.
Instruction Scope
The SKILL.md and read_config.py limit actions to scanning local config files (~/.openclaw/openclaw.json, ~/.openclaw/models.json, ~/.clawdbot/clawdbot.json, $OPENCLAW_CONFIG, and ./openclaw.json), checking env vars, and pinging localhost:11434 to detect Ollama. That scope is appropriate for a local router. Note: the README/SKILL.md mentions 'web-search-powered model library updates' but there is no code in the package that performs web searches or external lookups — this claim isn't implemented in the included script and should be treated as unimplemented or requiring additional code.
Install Mechanism
No install script or remote downloads are included; the skill is instruction-only with a small local Python script (no archives or remote URLs). This is low-risk from an installation perspective.
Credentials
The skill does not require any declared environment variables but deliberately checks for many provider API key names (both in parsed config files and in process environment). This is proportionate to its purpose (detecting which providers are activated). However, the skill reads local configuration files which may contain API keys or secrets; the provided script checks presence but does not print key values. Users should be aware the skill inspects files/ENV that can contain sensitive credentials.
Persistence & Privilege
The skill does not request always:true or other elevated persistence. It is user-invocable and does not modify other skills or system-wide settings. Autonomous invocation is permitted (platform default) but not combined with other red flags.
Assessment
This skill appears to do what it says: it reads your local OpenClaw config and environment to detect active providers and then maps available models into Premium/Balanced/Economy tiers. Before installing or running it: (1) Inspect your OpenClaw config files (~/.openclaw/openclaw.json, ~/.openclaw/models.json, ~/.clawdbot/clawdbot.json) for any stored API keys — consider removing secrets or using environment variables if you prefer. (2) Review scripts/read_config.py if you want to confirm it only checks presence of keys (it does not print key values) and only queries localhost:11434 for Ollama. (3) Note the SKILL.md mentions 'web-search-powered' updates but no web-update code is included — if you see an updated version that performs web requests, review the implementation. (4) Run the provided script locally (or in a sandbox) to verify behavior before granting it broader access.Like a lobster shell, security has layers — review code before you run it.
latestvk973e2e1zctfyfma758pvbnewh8396gw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🦞 Clawdis
Binspython3