MCP server that lets Claude Code delegate trivial, standalone questions to a local Gemma 4 model via llama.cpp, saving paid tokens.
Emerging. No concerning findings. Grades remain capped until the project builds reputation through adoption. Grades reflect the full trust pyramid: code, verification depth, and reputation. New projects cap at C until adoption is earned.
How we verified
Verified is a snapshot. Live keeps it current, and builds your track record.
⚡ Connect GitHub → continuous verification on every pushwhy connect →Who stands behind it
pakkio
Source: Glama
Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find. Or connect your repo for our deepest verification, Live Monitored: read-only, revoke anytime. What we access →
Install from
M8ven verifies MCPs across every public registry — install directly from whichever one you prefer.
LOCAL_LLM_CTX_SIZEcontext size passed to llama-cli for e2b/e4b (default 8192). The 12b entry hardcodes a smaller 4096 regardless of this var — see below.LOCAL_LLM_IDLE_TIMEOUT_MSLOCAL_LLM_MODELinitial active model, e2b, e4b, or 12b (or their aliases gemma-4-e2b/gemma-4-e4b/gemma-4-12b); default e2bLOCAL_LLM_NGLGPU layers to offload (default 99, i.e. all layers on GPU)LOCAL_LLM_PROJECT_DIRpath to the local_llm project (models/binaries live here; default /home/pakkio/w/local_llm)Tool annotations
No tools have read-only/destructive annotations
Add readOnlyHint or destructiveHint annotations to every tool so hosts can warn users before invoking.
All four hints declared on every tool
5/5 tools missing one or more hints — ask_local_llm (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); local_llm_health (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); switch_local_llm_model (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint), +2 more. OpenAI's directory rejects tools where any of the four hints are missing or non-boolean.
For every tool, set all four hints (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) to explicit true/false values that match the handler’s actual behaviour.
Tool inputs are validated
3/5 tool handlers declare input schemas (60%)
Declare an inputSchema with zod/joi/yup on every tool definition.
Tool handlers catch errors
3/5 tool handlers wrap calls in try/catch (60%)
Wrap each tool handler body in try/catch and return a structured error response.
License file
No license file
Add a LICENSE file (MIT, Apache-2.0, etc.).
Tests exist
No test files found
Add tests that exercise each declared tool.
Claim the listing to review these findings one by one and send us a correction where you disagree, straight to the team. Claiming also means we tell you when the grade moves, and reach you first if we find anything urgent.
[](https://m8ven.ai/mcp/pakkio-mcp-local-llm-1d9owd)?variant=verified to the badge URL.Vetting this one by hand? Tool Check is an MCP that scores other MCPs. Add it once and ask Claude, ChatGPT, or any MCP client to grade a server, surface CVEs, check the publisher, and suggest safer alternatives — before you install.
https://m8ven.ai/api/mcp/tool-check