A guide for running GGUF language models with llama.cpp, a tool for local AI inference. It helps choose model files and commands for CPUs and different kinds of GPUs.

Use it to discover GGUF models, choose Q4, Q5, Q6, or IQ variants, and build llama-server or llama-cli commands for local inference.

It makes local model setup less guesswork by matching quantized model versions to available RAM or VRAM. It can also help find compatible models on Hugging Face, a model-sharing site.

llama-cpp — Skill by john-data-chen

Install

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.

agentmods

npx agentmods add skills/john-data-chen/hermes-agent-backup/llama-cpp

Any agent

npx skills add john-data-chen/hermes-agent-backup --skill llama-cpp

Clone the repo

git clone --depth 1 https://github.com/john-data-chen/hermes-agent-backup

Made for: Claude Code, Codex.

Per session 18 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 2,512 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 1 finding. Scan, not verified.

Origin 98% copy Near-identical to another mod in the catalogue.

Token cost

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	$0.00018	$0.02512
Opus 5	$0.00009	$0.01256
Sonnet 5	$0.00004	$0.00502
Haiku 4.5	$0.00002	$0.00251

Measured yesterday against content hash 9c1ef45f6197, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

llama-cpp scanned grade A with 1 finding 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 yesterday.

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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl http://localhost:8080/v1/chat/completions \

Origin

This is a copy

98% identical to llama-cpp — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/mlops/inference/llama-cpp/SKILL.md · 250 lines

How it starts

The opening of the file, as written. The whole thing — 250 lines — stays where its author put it; the contents beside it link to each section on GitHub.

llama.cpp + GGUF

Use this skill for local GGUF inference, quant selection, or Hugging Face repo discovery for llama.cpp.

When to use

Run local models on CPU, Apple Silicon, CUDA, ROCm, or Intel GPUs
Find the right GGUF for a specific Hugging Face repo
Build a llama-server or llama-cli command from the Hub
Search the Hub for models that already support llama.cpp
Enumerate available .gguf files and sizes for a repo
Decide between Q4/Q5/Q6/IQ variants for the user's RAM or VRAM

Model Discovery workflow

Prefer URL workflows before asking for hf, Python, or custom scripts.

Search for candidate repos on the Hub:
- Base: https://huggingface.co/models?apps=llama.cpp&sort=trending
- Add search=<term> for a model family
- Add num_parameters=min:0,max:24B or similar when the user has size constraints
Open the repo with the llama.cpp local-app view:
- https://huggingface.co/<repo>?local-app=llama.cpp
Treat the local-app snippet as the source of truth when it is visible:
- copy the exact llama-server or llama-cli command
- report the recommended quant exactly as HF shows it
Read the same ?local-app=llama.cpp URL as page text or HTML and extract the section under Hardware compatibility:
- prefer its exact quant labels and sizes over generic tables
- keep repo-specific labels such as UD-Q4_K_M or IQ4_NL_XL
- if that section is not visible in the fetched page source, say so and fall back to the tree API plus generic quant guidance
Query the tree API to confirm what actually exists:
- https://huggingface.co/api/models/<repo>/tree/main?recursive=true
- keep entries where type is file and path ends with .gguf
- use path and size as the source of truth for filenames and byte sizes
- separate quantized checkpoints from mmproj-*.gguf projector files and BF16/ shard files
- use https://huggingface.co/<repo>/tree/main only as a human fallback
If the local-app snippet is not text-visible, reconstruct the command from the repo plus the chosen quant:
- shorthand quant selection: llama-server -hf <repo>:<QUANT>
- exact-file fallback: llama-server --hf-repo <repo> --hf-file <filename.gguf>
Only suggest conversion from Transformers weights if the repo does not already expose GGUF files.

Read the full file on GitHub · 250 lines

Files

What ships with it

6 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.

Changes

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.

yesterday First seen · 250 lines · 18 tokens per session scan A 9c1ef45f6197

Subscribe to this mod's changes

llama-cpp is a skill published in the GitHub repository john-data-chen/hermes-agent-backup (2 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 2,512 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). It is 98% identical to llama-cpp, differing in 2 lines, and is treated as a copy.

Data and AI agent agent-framework agent-skills agents ai ai-agent

create-agent

Creates and configures agent.yaml files, writes SOUL.md personality definitions, and sets up agent directory structures with skills, tools, and knowledge. Use when the user wants to configure an agent, create agent.yaml, write SOUL.md, set up agent directory structure, or customize agent settings.

open-gitagent/opengap · 62 tokens

run-agent

Configures and runs agents with different adapters including Claude, OpenAI, CrewAI, Lyzr, and GitHub Models. Supports local execution, remote git repos, and one-shot prompts. Use when the user wants to run an agent, switch LLM providers, configure adapter settings, or launch agents from git repositories.

open-gitagent/opengap · 68 tokens

regulatory-analysis

Analyzes documents and processes against FINRA, SEC, Federal Reserve, and CFPB regulatory frameworks. Identifies compliance gaps, classifies findings by severity, and recommends remediation. Use when performing compliance audits, regulatory reviews, gap analyses, or verifying policy adherence to financial regulations.

open-gitagent/opengap · 59 tokens

export-agent

Converts agent definitions between frameworks — exports to Claude Code, OpenAI, CrewAI, Lyzr, and GitHub Models formats, and imports from Claude, Cursor, and CrewAI projects. Use when the user wants to convert an agent, migrate to another framework, export to LangChain/AutoGen/CrewAI, or import from existing…

open-gitagent/opengap · 76 tokens

manage-skills

Searches the SkillsMP registry, installs skills locally or globally, creates custom skills with SKILL.md frontmatter, and manages the skill lifecycle. Use when the user wants to find skills, add new capabilities, install a skill, browse available skills, create a custom skill, or manage the skills system.

open-gitagent/opengap · 65 tokens

narco-check

Memory integrity audit. Detects hallucinations, circular confirmations, and state poisoning. Runs automatically after 2 consecutive failures or at nightly deep dive. Uses Opus 4.6 as the auditor model.

open-gitagent/opengap · 45 tokens

llama-cpp