How do I install the Llama Cpp Canister Release Test skill?

Run npx get-skill add tomevault-io/llama-cpp-canister-release-test in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Llama Cpp Canister Release Test skill do?

Test a llama_cpp_canister release zip end-to-end Use when this capability is needed. It is listed under Coding & Dev Tools on SkillMD.

Is Llama Cpp Canister Release Test safe to use?

SkillMD's automated safety review verdict for this skill is PASS. Independent scanners report: SkillSpector: WARNING, Skill Scanner: PASS. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Llama Cpp Canister Release Test?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Llama Cpp Canister Release Test free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Llama Cpp Canister Release Test?

tomevault-io (@tomevault-io) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Llama Cpp Canister Release Test by tomevault-io

Name: Llama Cpp Canister Release Test
Author: tomevault-io

File contents

Test llama_cpp_canister Release

Tests that a release zip contains everything needed and works end-to-end.

Follow these steps exactly in order. Abort and report on any failure.

1. Download & unzip release

Ask the user which release tag to test. If they say "latest", determine it with:

gh release view --repo onicai/llama_cpp_canister --json tagName --jq '.tagName'

Download and unzip:

rm -rf /tmp/llama_cpp_release_test
mkdir -p /tmp/llama_cpp_release_test
gh release download <TAG> --repo onicai/llama_cpp_canister --dir /tmp/llama_cpp_release_test
mkdir -p /tmp/llama_cpp_release_test/<TAG>
unzip /tmp/llama_cpp_release_test/llama_cpp_canister_<TAG>.zip -d /tmp/llama_cpp_release_test/<TAG>/

All subsequent commands run from /tmp/llama_cpp_release_test/<TAG>/.

2. Follow README setup instructions

Verify dfx

dfx --version

Verify the version is >= 0.31.0. If not, warn the user.

Create conda environment

source /opt/miniconda3/etc/profile.d/conda.sh
conda create -y -n llama_cpp_canister_release_test python=3.11
conda activate llama_cpp_canister_release_test

Install dependencies

cd /tmp/llama_cpp_release_test/<TAG>
pip install -r requirements.txt

3. Deploy

cd /tmp/llama_cpp_release_test/<TAG>
dfx start --clean --background
dfx deploy --network local
dfx ledger fabricate-cycles --canister llama_cpp --t 20

4. Download Qwen2.5 model from HuggingFace

cd /tmp/llama_cpp_release_test/<TAG>
mkdir -p models/Qwen/Qwen2.5-0.5B-Instruct-GGUF
wget -c -O models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf \
  https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF/resolve/main/qwen2.5-0.5b-instruct-q8_0.gguf

Verify SHA256:

echo "ca59ca7f13d0e15a8cfa77bd17e65d24f6844b554a7b6c12e07a5f89ff76844e  models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf" | shasum -a 256 -c

If the checksum does not match, abort immediately.

5. Upload model into canister

cd /tmp/llama_cpp_release_test/<TAG>
python -m scripts.upload --network local --canister llama_cpp \
  --canister-filename models/model.gguf --filetype gguf \
  --hf-sha256 "ca59ca7f13d0e15a8cfa77bd17e65d24f6844b554a7b6c12e07a5f89ff76844e" \
  models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf

6. Run pytest

cd /tmp/llama_cpp_release_test/<TAG>
pytest -vv --network local test/test_qwen2.py

If any tests fail, report the failures but continue to cleanup.

7. Cleanup

cd /tmp/llama_cpp_release_test/<TAG>
dfx stop

Ask the user if they want to remove the test directory and conda environment. If yes:

rm -rf /tmp/llama_cpp_release_test
source /opt/miniconda3/etc/profile.d/conda.sh
conda env remove -y -n llama_cpp_canister_release_test

8. Summary

Report:

Which release tag was tested
Whether all tests passed or which tests failed
Any issues encountered during the test

Source: onicai/llama_cpp_canister — distributed by TomeVault.

---name: llama-cpp-canister-release-testdescription: Test a llama_cpp_canister release zip end-to-end Use when this capability is needed.---# Test llama_cpp_canister ReleaseTests that a release zip contains everything needed and works end-to-end.Follow these steps exactly in order. Abort and report on any failure.## 1. Download & unzip releaseAsk the user which release tag to test. If they say "latest", determine it with:```bashgh release view --repo onicai/llama_cpp_canister --json tagName --jq '.tagName'```Download and unzip:```bashrm -rf /tmp/llama_cpp_release_testmkdir -p /tmp/llama_cpp_release_testgh release download <TAG> --repo onicai/llama_cpp_canister --dir /tmp/llama_cpp_release_testmkdir -p /tmp/llama_cpp_release_test/<TAG>unzip /tmp/llama_cpp_release_test/llama_cpp_canister_<TAG>.zip -d /tmp/llama_cpp_release_test/<TAG>/```All subsequent commands run from `/tmp/llama_cpp_release_test/<TAG>/`.## 2. Follow README setup instructions### Verify dfx```bashdfx --version```Verify the version is >= 0.31.0. If not, warn the user.### Create conda environment```bashsource /opt/miniconda3/etc/profile.d/conda.shconda create -y -n llama_cpp_canister_release_test python=3.11conda activate llama_cpp_canister_release_test```### Install dependencies```bashcd /tmp/llama_cpp_release_test/<TAG>pip install -r requirements.txt```## 3. Deploy```bashcd /tmp/llama_cpp_release_test/<TAG>dfx start --clean --backgrounddfx deploy --network localdfx ledger fabricate-cycles --canister llama_cpp --t 20```## 4. Download Qwen2.5 model from HuggingFace```bashcd /tmp/llama_cpp_release_test/<TAG>mkdir -p models/Qwen/Qwen2.5-0.5B-Instruct-GGUFwget -c -O models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf \ https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF/resolve/main/qwen2.5-0.5b-instruct-q8_0.gguf```Verify SHA256:```bashecho "ca59ca7f13d0e15a8cfa77bd17e65d24f6844b554a7b6c12e07a5f89ff76844e models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf" | shasum -a 256 -c```If the checksum does not match, abort immediately.## 5. Upload model into canister```bashcd /tmp/llama_cpp_release_test/<TAG>python -m scripts.upload --network local --canister llama_cpp \ --canister-filename models/model.gguf --filetype gguf \ --hf-sha256 "ca59ca7f13d0e15a8cfa77bd17e65d24f6844b554a7b6c12e07a5f89ff76844e" \ models/Qwen/Qwen2.5-0.5B-Instruct-GGUF/qwen2.5-0.5b-instruct-q8_0.gguf```## 6. Run pytest```bashcd /tmp/llama_cpp_release_test/<TAG>pytest -vv --network local test/test_qwen2.py```If any tests fail, report the failures but continue to cleanup.## 7. Cleanup```bashcd /tmp/llama_cpp_release_test/<TAG>dfx stop```Ask the user if they want to remove the test directory and conda environment. If yes:```bashrm -rf /tmp/llama_cpp_release_testsource /opt/miniconda3/etc/profile.d/conda.shconda env remove -y -n llama_cpp_canister_release_test```## 8. SummaryReport:- Which release tag was tested- Whether all tests passed or which tests failed- Any issues encountered during the test---> Source: [onicai/llama_cpp_canister](https://github.com/onicai/llama_cpp_canister) — distributed by [TomeVault](https://tomevault.io).

Frequently asked questions

Llama Cpp Canister Release Test