ClawHub

Deepseek Deepseek Coder

v1.0.3

DeepSeek DeepSeek-Coder — run DeepSeek-V3, DeepSeek-R1, DeepSeek-Coder across your local fleet. 7-signal scoring routes every request to the best device. Cro...

⭐ 2· 71·1 current·1 all-time
byTwin Geeks@twinsgeeks
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Pending
View report →
OpenClawOpenClaw
Benign
medium confidence
✓
Purpose & Capability
Name/description, examples (curl/OpenAI SDK), and required tools (curl/wget, optional python/pip) all align with running a local fleet router and calling a localhost Ollama-compatible API. The referenced config paths (~/.fleet-manager/latency.db, ~/.fleet-manager/logs/herd.jsonl) are consistent with a fleet manager's state and logs.
✓
Instruction Scope
SKILL.md instructs installing 'ollama-herd' and running local processes (herd, herd-node) and making requests to localhost:11435. It does not instruct reading unrelated system files, exporting secrets, or sending data to external endpoints other than pulling models on demand (which the docs say requires confirmation).
ℹ
Install Mechanism
This is instruction-only with no install spec; the docs instruct users to 'pip install ollama-herd'. That is coherent but introduces typical supply-chain risk because installing a PyPI package runs third-party code on your machine and may trigger on-demand model downloads. The SKILL itself does not include or pin any binaries.
✓
Credentials
No environment variables or credentials are requested. Example code uses a localhost base_url and sets api_key to 'not-needed'. There are no unexpected credential requests in the SKILL.md or metadata.
✓
Persistence & Privilege
always is false (not force-included). The skill does not request elevated platform privileges or attempt to modify other skills' configuration. Autonomous invocation is allowed by default, which is normal; nothing else indicates persistent privileged presence.
Assessment
This skill is internally consistent with its purpose, but before installing: (1) verify the PyPI package (ollama-herd) and the GitHub repo (review recent commits, maintainer identity, and issues); (2) install in a virtualenv or isolated VM if you are unsure; (3) be prepared for very large model downloads and ensure disk space and bandwidth; (4) confirm any model downloads prompt you before proceeding (the docs claim confirmation is required); (5) check what filesystem paths the herd service writes to (~/.fleet-manager/...) and restrict permissions if needed; (6) consider running the router behind a firewall or localhost-only interface to avoid exposing models to the network. If you want higher confidence, request the actual repository code or a pinned package artifact for review before installing.

Like a lobster shell, security has layers — review code before you run it.

apple-siliconvk97224zp6f4ma7p5zs7f1mz7n983ys80code-generationvk976xn4crebhdwyh3ngr8e5wan83xgt6codingvk97224zp6f4ma7p5zs7f1mz7n983ys80deepseekvk97224zp6f4ma7p5zs7f1mz7n983ys80deepseek-codervk97224zp6f4ma7p5zs7f1mz7n983ys80deepseek-r1vk97224zp6f4ma7p5zs7f1mz7n983ys80deepseek-v3vk97224zp6f4ma7p5zs7f1mz7n983ys80fleet-routingvk97224zp6f4ma7p5zs7f1mz7n983ys80latestvk972jpg12519djm4pq9zqqy6tx844d7ylocal-llmvk97224zp6f4ma7p5zs7f1mz7n983ys80ollamavk97224zp6f4ma7p5zs7f1mz7n983ys80reasoningvk97224zp6f4ma7p5zs7f1mz7n983ys80self-hostedvk976xn4crebhdwyh3ngr8e5wan83xgt6

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Runtime requirements

brain Clawdis
OSmacOS · Linux · Windows
Any bincurl, wget

Comments