MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. Each candidate is checked by a language model from its name, description and topics, and only actual MCP servers are listed (not clients, SDKs, lists or skills). Package metadata comes from ecosyste.ms (CC BY-SA 4.0). A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
@yun520-1/agent-trust-layergithub.com/yun520-1/agent-trust-layerupdated this year
Agent Trust Layer — rule-based discriminator engine for AI agent output verification. 47 dimensions, 9 check layers, 131 MCP tools. Zero LLM dependency. Checks what AI says before it reaches humans and says "no" when something is wrong.
npm MIT -
Free MCP server: score any AI agent output 0-100 and get a PASS/PARTIAL/FAIL verdict (VOP verification). Verification only, no money moves.
MIT -
steplockgithub.com/l0kifs/steplock-ai-skill-executorupdated this year
An MCP server gatekeeper runtime that enforces step-by-step AI skill execution with output verification, ensuring stable and reliable agent workflows regardless of LLM compliance.
pypi MIT License Copyright (c) 2026 Sergei Konovalov -
Local-first memory + output verification for AI agents. Zero runtime dependencies, works offline.
pypi AGPL-3.0 -
MCP server for verifying AI agent claims vs reality — inline grounding-check on retrieval context, swallowed-exception scanner on agent-written code, and multi-turn transcript review for unverified completions + cross-turn contradictions. Sub-second, local, free, MCP-native — the lightweight complement to dashboard-based eval frameworks (DeepEval, Phoenix, LangSmith).
pypi MIT License -
Verify before you trust. Cross-provider verification with LLMs and Jev. CLI, library & MCP.
npm MIT -
A standalone MCP suppressor that filters, stabilises, and sanitises tool responses to prevent hallucinated or unsafe output.
NOASSERTION -
A deterministic first-pass filter for unsupported claims in LLM output, computed without calling another LLM. Not a standalone hallucination detector. A spec plus reference implementation, shipped as touchstone-mcp.
pypi Apache-2.0 -
mcp-grounding-enforcergithub.com/steveswain14/mcp-grounding-enforcerupdated this year
A standalone MCP suppressor that enforces grounding by validating references, rejecting unsupported claims, and stabilising model outputs.
NOASSERTION -
@certnode/mcp-serverupdated this year
CertNode MCP Server — verify content authenticity, sign AI output for EU AI Act / FRE 902 compliance, track AI code generation accountability
npm MIT -
dlx-public-kitgithub.com/kodomonocch1/dlx-public-kitupdated this year
# DLX Public Kit DLX Phase 1 is a public evaluator kit for a thin MCP-native trust proxy. The public face is intentionally narrow: evaluators call `extract`, then `get_execution`, and judge the trust boundary by whether work ends in `succeeded` or `failed_safe`. The gateway remains behind that surface as an internal / secondary delivery layer. It
★ 1