New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Llama Patch Changes

by Mesh-LLM in Mesh-LLM/mesh-llm

A skill your agent uses when changing mesh-llm's llama.cpp patch queue, upstream pin, prepare/build scripts, or carried RPC, MoE, and mesh-hook llama.cpp patches.

Apache-2.0Auto-check passedAI & LLM Engineering

Install Llama Patch Changes

skills CLI
$ npx skills add Mesh-LLM/mesh-llm --skill llama-patch-changes -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install Mesh-LLM/mesh-llm llama-patch-changes --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/Mesh-LLM/mesh-llm.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/llama-patch-changes .claude/skills/llama-patch-changes && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
llama-patch-changes
GitHub stars
3.5k
Token cost
~1.9k tokens
SKILL.md length
712 words
Files
1
Skills in repo
25
Repo updated
First seen
Licence
Apache-2.0

At a glance

A skill your agent uses when changing mesh-llm's llama.cpp patch queue, upstream pin, prepare/build scripts, or carried RPC, MoE, and mesh-hook llama.cpp patches.

  • Changing mesh-llms llama.cpp patch queue
  • SKILL.md covers Boundaries, Local Flow, Validation and Updating The Upstream Pin, plus 1 more section
  • Calls git, cargo and just
  • Prepare/build scripts

What it does

Llama Patch Changes is an agent skill from Mesh-LLM/mesh-llm. Use when changing mesh-llm's llama.cpp patch queue, upstream pin, prepare/build scripts, or carried RPC, MoE, and mesh-hook llama.cpp patches.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in AI & LLM Engineering, covering LLM inference and serving. It works with llama.cpp. The repository describes itself as: Distributed AI/LLM for the people. Share compute privately or publicly to power your agents and chat. The licence is Apache-2.0.

When your agent uses it

  • Changing mesh-llms llama.cpp patch queue
  • Prepare/build scripts
  • Mesh-hook llama.cpp patches

Example prompts

  • “/llama-patch-changes”

What it can do on your machine

Read from SKILL.md and the folder at commit 48bf685. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • cargo
    • just
    • curl

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git and curl, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Llama Patch Changes loads about 1.9k tokens when it runs. Until then it costs about 41 tokens; SKILL.md has 712 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~41
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Mesh-LLM/mesh-llm at commit 48bf685, republished under its Apache-2.0 licence (© Mesh-LLM). 712 words, ~1,882 tokens.

Download SKILL.mdSave it as .claude/skills/llama-patch-changes/SKILL.md (or your agent's skills folder).
name
llama-patch-changes
description
Use when changing mesh-llm's llama.cpp patch queue, upstream pin, prepare/build scripts, or carried RPC, MoE, and mesh-hook llama.cpp patches.

llama-patch-changes

Use this skill when editing the llama.cpp patch queue, refreshing patches from a llama.cpp checkout, updating the pinned upstream SHA, or changing build scripts that prepare or consume patched llama.cpp.

Boundaries

  • Keep durable llama-side changes in the ordered queue under skippy/llama_cpp/patches: top-level core patches first, model_support/series second, and generated/series last.
  • Keep the upstream pin in skippy/llama_cpp/upstream.txt.
  • Do not add a submodule, vendor a llama checkout, or depend on the old Mesh-LLM llama.cpp fork.
  • Do not treat edits in .deps/llama.cpp as durable until the patch queue has been regenerated and committed.
  • Do not add llama-stage ABI/static in-process patches unless the task explicitly asks for that integration pass.
  • Prefer small, reviewable llama commits with one functional boundary per patch. Keep patch numbers unique and contiguous within each queue lane.
  • Put a new model family's implementation, conversion, templates, multimodal integration, runtime adaptations, and family tests in one focused patch in model_support/. Do not spread family-specific code through core patches.
  • Keep generated graph-semantics edits in generated/; do not hand-maintain them in either the core or model-support lane.
  • Do not append a terminal patch whose only purpose is to split, move, or clean up code introduced by earlier patches. Recreate the affected patches so they use the intended ownership boundaries from the outset.

Local Flow

Prepare the pinned upstream checkout and current patch queue:

bash
scripts/prepare-llama.sh pinned

For actual llama-side editing, prefer a normal llama.cpp checkout or branch where commits can be named and inspected. Base the branch on upstream ggml-org/llama.cpp master, then carry the Mesh-LLM patch commits on top.

For a deliberate queue rewrite, reconstruct capability-owned core commits from the pinned upstream, add model-family support commits, then add the generated family shards. Verify the reconstructed head is tree-identical to the authoritative final checkout before regenerating each queue lane. Preserve the model_support/series and generated/series manifests explicitly rather than flattening their patches into the top-level queue.

bash
repo_root="$(pwd)"
llama_checkout="${LLAMA_CHECKOUT:-$repo_root/.deps/llama.cpp}"
patch_backup="$(mktemp -d /tmp/mesh-llm-patches.XXXXXX)"
patch_root="$repo_root/skippy/llama_cpp/patches"
mkdir -p "$patch_backup/core"
mv "$patch_root"/*.patch "$patch_backup/core/"
mv "$patch_root/model_support" "$patch_backup/model_support"
mkdir -p "$patch_root/model_support"
git -C "$llama_checkout" format-patch \
  --start-number 1 \
  --output-directory "$patch_root" \
  "$(cat "$repo_root/skippy/llama_cpp/upstream.txt")..<core-head>"
git -C "$llama_checkout" format-patch \
  --start-number 1 \
  --output-directory "$patch_root/model_support" \
  "<core-head>..<model-support-head>"

Regenerate model_support/series from the sorted patch filenames. Leave the generated lane in place unless its generator inputs changed; if they did, regenerate it with the deterministic family-patch workflow rather than moving or formatting those commits by hand.

Keep the temporary backup until clean patch application and the required native build pass. Ordinary focused changes may append a patch without rebuilding unrelated functional boundaries.

Validation

Validate that patches apply in a clean checkout:

bash
tmp_llama="$(mktemp -d /tmp/mesh-llm-llama.XXXXXX)"
trap 'rm -rf -- "$tmp_llama"' EXIT
LLAMA_WORKDIR="$tmp_llama" scripts/prepare-llama.sh pinned

For normal mesh-llm validation, use the repository build workflow:

bash
just build

For Rust-only fallout from build-system or runtime call-site changes:

bash
cargo fmt --all --check
cargo check -p mesh-llm

Run Cargo commands serially. This repo frequently hits Cargo lock conflicts when multiple Cargo commands run at once.

Show full SKILL.md (296 more words)Show less
Model-load spot check (required for backend or model-switch changes)

A clean patch replay plus a green build does not prove the runtime works. Metal shaders in ggml-metal.metal are JIT-compiled on-device at first model open, so a broken shader builds green everywhere and only fails at load time. Machine-reconciled patches can also silently drop arch cases from switches in src/llama-model.cpp (for example llama_model_rope_type), which only fail when a model of that arch creates a context.

After any queue change that touches backend sources (.metal, CUDA, Vulkan), ggml.c/ggml-*.h kernel argument structs, or src/llama-model.cpp switch statements, load a small real model on your local backend and confirm one completion returns:

bash
./target/debug/mesh-llm serve --model "Qwen/Qwen2.5-3B-Instruct-GGUF@main:q4_k_m" --log-format json
# wait for the model to appear, then:
curl -s http://127.0.0.1:9337/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"Qwen/Qwen2.5-3B-Instruct-GGUF:q4_k_m","messages":[{"role":"user","content":"Say OK"}],"max_tokens":5}'

On a Mac this exercises the Metal shader JIT path directly. Watch the JSON log for metal_library_init: error and model-open failure. If the change adds or touches support for a specific model family, spot-check a model of that family as well.

When regenerating the queue against a new upstream pin, also diff the arch case lists of reconciled switches against upstream and account for every deletion:

bash
# example: rope-type switch parity
git -C .deps/llama.cpp show <upstream-pin>:src/llama-model.cpp \
  | awk '/llama_rope_type llama_model_rope_type/,/^}$/' \
  | grep -oE "LLM_ARCH_[A-Z0-9_]+" | sort > /tmp/upstream-cases.txt
awk '/llama_rope_type llama_model_rope_type/,/^}$/' .deps/llama.cpp/src/llama-model.cpp \
  | grep -oE "LLM_ARCH_[A-Z0-9_]+" | sort > /tmp/patched-cases.txt
comm -23 /tmp/upstream-cases.txt /tmp/patched-cases.txt  # must be empty or explained

Updating The Upstream Pin

Test the queue against current upstream without moving the pin:

bash
scripts/prepare-llama.sh latest
just build
cargo test -p mesh-llm --lib

If the queue applies and validation passes, update the upstream pin:

bash
cp skippy/llama_cpp/upstream.txt /tmp/old-llama-upstream.txt
git -C .deps/llama.cpp rev-parse "$(cat .deps/llama.cpp/.git/mesh-llm-upstream-sha)" > skippy/llama_cpp/upstream.txt

Commit the pin update with any patch refreshes.

Gotchas

  • scripts/prepare-llama.sh configures local git identity for git am; keep that responsibility there for fresh CI checkouts.
  • Patch files are mail-format artifacts and may intentionally contain whitespace that git diff --check reports. Do not hand-normalize patches in a way that changes or breaks git am.
  • Build outputs live under .deps/llama.cpp/build; the root llama.cpp symlink is compatibility-only.
  • Important backend flags include GGML_RPC=ON, BUILD_SHARED_LIBS=OFF, and LLAMA_OPENSSL=OFF; preserve CPU, Metal, CUDA, Vulkan, and ROCm behavior when touching build scripts.
  • See mesh-llm/docs/LLAMA_CPP_FORK.md for the full patch-queue maintenance notes and mesh-llm/docs/LLAMA_STAGE_INTEGRATION_PLAN.md for deferred llama-stage integration.

© Mesh-LLM, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/llama-patch-changes of Mesh-LLM/mesh-llm.

Open the folder on GitHubat commit 48bf685

Compare with similar skills

Llama Patch Changes next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Llama Patch Changes compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Llama Patch Changes this skillMesh-LLM/mesh-llm3.5k—~1.9kAutomated safety check: PassApache-2.0
Qwen Mtp GgufR6410418/Jackrong-llm-finetuning-guide1.7k—~1.7kAutomated safety check: PassMIT
Hugging Face Local Modelshuggingface/skills11k3 repos~945Automated safety check: PassApache-2.0
Add Quantization Datatypeintel/auto-round1.6k—~1.5kAutomated safety check: PassApache-2.0
Quantized Exportwshobson/agents40k—~2kAutomated safety check: PassMIT
Distil Pii RedactorHybridAIOne/hybridclaw158—~1kAutomated safety check: PassMIT

Similar skills

  • Qwen Mtp Gguf

    R6410418/Jackrong-llm-finetuning-guide

    Complete agent-ready workflow for Qwen-family MTP or nextn GGUF conversion and release.

    1.7k GitHub stars~1.7k tokensUpdated 2 mo ago
    AI & LLM EngineeringAuto-check passed

  • Hugging Face Local Models

    huggingface/skills

    Official

    Finds llama.cpp-compatible GGUF models on the Hugging Face Hub, picks a quantization for your hardware and launches them with llama-cli or llama-server.

    11k GitHub starsUsed in 3 repos~945 tokens
    AI & LLM EngineeringAuto-check passed

  • Add Quantization Datatype

    intel/auto-round

    Official

    Add a new quantization data type to AutoRound (e.g., INT, FP8, MXFP, NVFP, GGUF variants).

    1.6k GitHub stars~1.5k tokensUpdated today
    AI & LLM EngineeringAuto-check passed

  • Quantized Export

    wshobson/agents

    Export a promoted fine-tuned model in the right deployment format — merged safetensors, LoRA-only, GGUF with imatrix, or FP8.

    40k GitHub stars~2k tokensUpdated 3 days ago
    AI & LLM EngineeringAuto-check passed

  • Distil Pii Redactor

    HybridAIOne/hybridclaw

    Redact, anonymize, sanitize, or remove PII locally with Distil-PII and llama.cpp; keep personal data and secret values out of model context, logs, and chat.

    158 GitHub stars~1k tokensUpdated today
    AI & LLM EngineeringAuto-check passed

  • ML Research Lab

    AnastasiyaW/codex-claude-code-config

    Machine-learning research loop for dataset curation, fine-tuning, evaluation, inference deployment, experiment tracking, and model explainability.

    154 GitHub stars~794 tokensUpdated today
    AI & LLM EngineeringAuto-check passed

More from Mesh-LLM/mesh-llm

All 25 skills in this repo

  • Release Validation

    Mesh-LLM/mesh-llm

    A skill your agent uses when validating a MeshLLM release candidate or current HEAD against the last GitHub release, assembling the canonical feature/fix/modification inventory, testing locally…

    3.5k GitHub stars~2.6k tokensUpdated today
    Auto-check passed

  • Benchmark Tune

    Mesh-LLM/mesh-llm

    A skill your agent uses when running, debugging, interpreting, or documenting mesh-llm benchmark tune model-serving throughput trials, including choosing…

    3.5k GitHub stars~1.6k tokensUpdated today
    Auto-check passed

  • Config Settings Management

    Mesh-LLM/mesh-llm

    A skill your agent uses when adding, renaming, removing, validating, or exposing mesh-llm config settings, including built-in settings, plugin config schemas, owner-control apply behavior, CLI…

    3.5k GitHub stars~1.4k tokensUpdated today
    Auto-check passed

  • Connect Agents

    Mesh-LLM/mesh-llm

    A skill your agent uses when connecting agent tools or OpenAI clients to mesh-llm — launching or configuring Goose, Claude Code, OpenCode, Pi, curl, or any OpenAI-compatible client against a local…

    3.5k GitHub stars~1.2k tokensUpdated today
    Auto-check passed

  • Hf Bf16 Gguf Conversion Jobs

    Mesh-LLM/mesh-llm

    A skill your agent uses when converting Hugging Face SafeTensors checkpoints into split BF16 GGUF model repos with skippy-quantize on Hugging Face Jobs or a local machine, then publishing the…

    3.5k GitHub stars~1.1k tokensUpdated today
    Auto-check passed

  • Hf Gguf Quant Jobs

    Mesh-LLM/mesh-llm

    A skill your agent uses when creating, monitoring, validating, or documenting low-memory Hugging Face Jobs or local runs that quantize split BF16/FP16 GGUF model repos into custom quant GGUF repos…

    3.5k GitHub stars~1.9k tokensUpdated today
    Auto-check passed

Works with

Categories

Questions about Llama Patch Changes

What does Llama Patch Changes do?

A skill your agent uses when changing mesh-llm's llama.cpp patch queue, upstream pin, prepare/build scripts, or carried RPC, MoE, and mesh-hook llama.cpp patches. Llama Patch Changes is an agent skill from Mesh-LLM/mesh-llm.cpp patches.

When should I use Llama Patch Changes?

Llama Patch Changes fits situations like: changing mesh-llms llama.cpp patch queue; prepare/build scripts; mesh-hook llama.cpp patches.

How do I install Llama Patch Changes in Claude Code?

Run `npx skills add Mesh-LLM/mesh-llm --skill llama-patch-changes -a claude-code`. Or copy the skill folder (.agents/skills/llama-patch-changes in Mesh-LLM/mesh-llm) into .claude/skills/llama-patch-changes in your project. Claude Code loads it when a task matches its description.

How do I install Llama Patch Changes in Codex?

Run `npx skills add Mesh-LLM/mesh-llm --skill llama-patch-changes -a codex`. Or copy the skill folder (.agents/skills/llama-patch-changes in Mesh-LLM/mesh-llm) into .agents/skills/llama-patch-changes in your project. Codex loads it when a task matches its description.

Can I use Llama Patch Changes in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Mesh-LLM/mesh-llm --skill llama-patch-changes -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/llama-patch-changes, .gemini/skills/llama-patch-changes, .github/skills/llama-patch-changes and .opencode/skills/llama-patch-changes in your project.

What does Llama Patch Changes need to run?

Going by SKILL.md and its folder, Llama Patch Changes needs the command-line tools its instructions call (git, cargo, just and curl).

Does Llama Patch Changes access the network?

SKILL.md contains no URLs. Its commands use git and curl, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Llama Patch Changes safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Llama Patch Changes use?

Llama Patch Changes is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Llama Patch Changes use?

About 1.9k tokens (SKILL.md is roughly 7.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Llama Patch Changes?

Skills that share tags, products or a category with Llama Patch Changes: Qwen Mtp Gguf (R6410418/Jackrong-llm-finetuning-guide, 1.7k stars), Hugging Face Local Models (huggingface/skills, 11k stars), Add Quantization Datatype (intel/auto-round, 1.6k stars) and Quantized Export (wshobson/agents, 40k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Llama Patch Changes?

Mesh-LLM (a GitHub organization) maintains it in Mesh-LLM/mesh-llm, which has 3,485 GitHub stars. The repository holds 25 skills in this directory. The repository was last updated on October 8, 2026.

Source: Mesh-LLM/mesh-llm on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.