Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ArieGoldkin/claude-forge --skill ollama-localgit clone --depth 1 https://github.com/ArieGoldkin/claude-forgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ariegoldkin/claude-forge/ollama-local)<a href="https://agentmods.dev/skills/ariegoldkin/claude-forge/ollama-local"><img src="https://agentmods.dev/badge/skills/ariegoldkin/claude-forge/ollama-local/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ariegoldkin/claude-forge/ollama-local"><img src="https://agentmods.dev/badge/skills/ariegoldkin/claude-forge/ollama-local.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00040 | $0.01334 |
| Opus 5 | $0.00020 | $0.00667 |
| Sonnet 5 | $0.00008 | $0.00267 |
| Haiku 4.5 | $0.00004 | $0.00133 |
Grade C, and why
ollama-local scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -fsSL https://ollama.ai/install.sh | sh Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsSL https://ollama.ai/install.sh | sh How it starts
The opening of the file, as written. The whole thing — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ollama Local Inference
Run LLMs locally for cost savings, privacy, and offline development.
Quick Start
# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh
# Pull models
ollama pull deepseek-r1:70b # Reasoning (GPT-4 level)
ollama pull qwen2.5-coder:32b # Coding
ollama pull nomic-embed-text # Embeddings
# Start server
ollama serve
Recommended Models (M4 Max 256GB)
| Task | Model | Size | Notes |
|---|---|---|---|
| Reasoning | deepseek-r1:70b |
~42GB | GPT-4 level |
| Coding | qwen2.5-coder:32b |
~35GB | 73.7% Aider benchmark |
| Embeddings | nomic-embed-text |
~0.5GB | 768 dims, fast |
| General | llama3.2:70b |
~40GB | Good all-around |
LangChain Integration
from langchain_ollama import ChatOllama, OllamaEmbeddings
# Chat model
llm = ChatOllama(
model="deepseek-r1:70b",
base_url="http://localhost:11434",
temperature=0.0,
num_ctx=32768, # Context window
keep_alive="5m", # Keep model loaded
)
# Embeddings
embeddings = OllamaEmbeddings(
model="nomic-embed-text",
base_url="http://localhost:11434",
)
# Generate
response = await llm.ainvoke("Explain async/await")
vector = await embeddings.aembed_query("search text")
Tool Calling with Ollama
from langchain_core.tools import tool
@tool
def search_docs(query: str) -> str:
"""Search the document database."""
return f"Found results for: {query}"
# Bind tools
llm_with_tools = llm.bind_tools([search_docs])
response = await llm_with_tools.ainvoke("Search for Python patterns")
Structured Output
from pydantic import BaseModel, Field
class CodeAnalysis(BaseModel):
language: str = Field(description="Programming language")
complexity: int = Field(ge=1, le=10)
issues: list[str] = Field(description="Found issues")
structured_llm = llm.with_structured_output(CodeAnalysis)
result = await structured_llm.ainvoke("Analyze this code: ...")
# result is typed CodeAnalysis object
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 190 lines · 40 tokens per session scan C bb3aeb798daf
ollama-local is a skill published in the GitHub repository ArieGoldkin/claude-forge (6 stars, last pushed 1mo ago), licensed MIT. It adds 40 tokens to every session and 1,334 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-memory-thermal-ops
Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
langchain
Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, Anthropic, Google), 500+ integrations, ReAct agents, tool calling, memory management, and vector store retrieval. Use for building chatbots, question-answering systems, autonomous agents, or RAG…
llamaindex
Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing, and querying. Features vector indices, query engines, agents, and multi-modal support. Use for document Q&A, chatbots, knowledge retrieval, or building RAG pipelines. Best for data-centric LLM…
sglang
Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic workflows with tool calls, or when you need 5× faster inference than vLLM with prefix sharing. Powers 300,000+ GPUs at xAI, AMD, NVIDIA, and LinkedIn.