minicpm5-deploy-llama-cpp

Run MiniCPM5-1B or MiniCPM5-2B with llama.cpp using the released GGUF artifacts (F16 / Q8_0 / Q4_K_M). Use when the user wants CPU-only / consumer-GPU / cross-platform native deployment, asks for "llama.cpp", "llama-cli", "llama-server", "GGUF", or has no Python available.

Quality

95%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

90%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, executable deployment skill with excellent conciseness and actionability and a clear validated workflow. Its only gaps are a missing fix-retry feedback loop around the validation step and a dead external reference path.

Suggestions

Add an explicit error-recovery loop after the Validate step (e.g., 'If the reply is not "2", re-check -ngl/-c values and the GGUF path, then rerun llama-server').

Fix or remove the broken Reference link to docs/deployment/llama_cpp.md, or replace it with an inline summary so navigation is not dead-ended.

Consider a brief one-line 'Expected' check for the CLI path (step 3a) so both run modes have validation, matching step 4.

Dimension	Reasoning	Score
Conciseness	Lean tables and copy-paste commands with almost no padding; the one explanatory paragraph (the min_p=0.0 repetition-loop rationale) earns its place as non-obvious pitfall knowledge rather than restating basics Claude already knows.	5 / 5
Actionability	Fully executable install, download, serve, and validate commands with a concrete defaults table (GGUF_REPO/QUANT/NGL/CTX); specific examples cover the common CLI and HTTP-server cases.	5 / 5
Workflow Clarity	Clear numbered sequence (1-4) with an explicit Validate step and an expected-output check ('Expected: "2" in the reply'), but it lacks an explicit error-recovery feedback loop (fix-then-revalidate) for the validation step.	4 / 5
Progressive Disclosure	Well-organized sections with clearly signaled one-level references ('When NOT to use' sibling pointers, LoRA cross-reference to another skill), but the lone external doc link (docs/deployment/llama_cpp.md) does not resolve to a real file, slightly degrading navigation.	4 / 5
	Total	18 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: third-person, concrete actions, explicit 'Use when' trigger clause with natural keywords, and a clear niche that distinguishes it from sibling deployment skills. No significant weaknesses.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions across deployment targets ('Run MiniCPM5-1B or MiniCPM5-2B with llama.cpp using the released GGUF artifacts (F16 / Q8_0 / Q4_K_M)', 'CPU-only / consumer-GPU / cross-platform native deployment') with comprehensive coverage.	5 / 5
Completeness	Explicitly answers both what ('Run MiniCPM5... with llama.cpp using the released GGUF artifacts') and when ('Use when the user wants CPU-only / consumer-GPU... asks for "llama.cpp"... or has no Python available') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive coverage of natural terms users say: '"llama.cpp", "llama-cli", "llama-server", "GGUF"' plus model names and deployment scenarios, including exact tool-name synonyms.	5 / 5
Distinctiveness Conflict Risk	Clear niche (MiniCPM5 + llama.cpp GGUF native deployment) with distinct triggers and minimal conflict risk versus sibling skills (vllm/mlx/ollama/lmstudio) named in the body.	5 / 5
	Total	20 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 1 suspicious	Warning

	Total	15 / 16 Passed

Repository: OpenBMB/MiniCPM
Path: skills/minicpm5-deploy-llama-cpp/SKILL.md
Commit: 316cfb1

Reviewed: 15 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.