llmfit
v0.2.2Detect local hardware (RAM, CPU, GPU/VRAM) and recommend the best-fit local LLM models with optimal quantization, speed estimates, and fit scoring.
⭐ 1· 898·7 current·7 all-time
byAlex Jones@alexsjones
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name and description match the runtime instructions: the SKILL.md tells the agent to run the llmfit CLI to detect hardware and produce model recommendations. Required binaries (llmfit) are consistent with that purpose and no unrelated credentials or files are requested.
Instruction Scope
Instructions are narrowly scoped to running llmfit commands (system, recommend) and mapping outputs to local providers (Ollama, vLLM, LM Studio). They do not instruct reading arbitrary system files or exfiltrating secrets. The skill suggests editing openclaw.json to configure models, which is coherent with its goal.
Install Mechanism
Install metadata is inconsistent and potentially risky: the SKILL.md / registry lists a Homebrew formula 'AlexsJones/llmfit' (a third‑party tap) and a second install entry that is labeled as 'cargo install llmfit' but is marked kind: 'node' (and registry also lists 'node'). This mismatch (node vs cargo label) is sloppy and prevents clear vetting of the install source. Homebrew taps from unknown owners should be reviewed before use because they install binaries from third parties.
Credentials
The skill requests no environment variables or credentials. That is proportionate for a local hardware-detection and recommendation tool.
Persistence & Privilege
always is false and the skill does not request or auto-modify other skills' configs. It only recommends edits to openclaw.json (user-driven). No privileged or persistent presence is requested by the skill metadata or instructions.
What to consider before installing
This skill appears to do what it says (run the llmfit CLI and recommend models), but the install metadata is inconsistent and the Homebrew tap is a third-party source. Before installing or running anything: 1) ask the maintainer for the upstream source or GitHub repo and a homepage; 2) inspect the Homebrew formula or cargo package code on a trusted repo; 3) prefer installing from an official, traceable release (GitHub releases or crates.io) rather than an anonymous tap; 4) if you must run the binary first, run `llmfit --version` and `llmfit --json system` in a sandbox/container to inspect output; 5) do not provide credentials or elevated privileges to this tool. If you want help vetting the brew formula or package repository, provide the install URL and I can point out risky patterns.Like a lobster shell, security has layers — review code before you run it.
latestvk979mvm0m95ka4j8fgh5h5z9zs819h3r
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🧠 Clawdis
Binsllmfit
Install
Install llmfit (brew tap AlexsJones/llmfit && brew install llmfit)
Bins: llmfit
brew install AlexsJones/llmfitInstall llmfit (cargo install llmfit)
Bins: llmfit