Developer Tools MCP Servers
Code, Git, APIs, SDKs, CI/CD, developer workflows, and local developer tooling.
Autorank Mcp
Published by GitHub account RyanKramer
Local MCP server of offline technical SEO utilities that let an assistant generate and validate structured data, meta tags, robots directives and SERP previews.
Agent Harnesses
Published by GitHub account RyanAlberts
Curated directory of AI agent harnesses: lets an assistant recommend, compare and look up harnesses, guides, templates and skills.
Runeward
Published by GitHub account Runewardd
Runeward (Rune Box) puts policy, approvals, budgets, sandboxed execution and signed audit evidence around an AI agent's tool calls.
Gatekeeper
Published by GitHub account Runestone-Labs
A local MCP front door that routes an assistant's shell, file-write and HTTP calls through Runestone Gatekeeper for policy checks, human approval and audit.
Agentpay Gateway Mcp
Published by GitHub account Rumblingb
A self-hosted gateway that routes an assistant's tool calls to many backend MCP servers, with credit-based per-call billing behind one API key.
Agentpay Feed Mcp
Published by GitHub account Rumblingb
A local MCP server that gives an assistant a real-time newsfeed of newly published MCP tools, capabilities and market events, with optional publishing.
Assay MCP Server
Published by GitHub account Rul1an
A policy-enforcing MCP proxy that allows or denies privileged tool calls before forwarding and records tamper-evident, offline-verifiable evidence.
Assay
Published by GitHub account Rul1an
A local MCP proxy that enforces allow/deny policy on tool calls and records tamper-evident, offline-verifiable evidence bundles for review.
MySQL Legacy MCP
Published by GitHub account Rufflet
Local stdio MCP server that lets an assistant inspect and query legacy MySQL 5.0–5.6 databases, read-only by default with opt-in write tools.
Toolgovern
Published by GitHub account RudrenduPaul
Wraps the toolgovern CLI as one generic MCP tool so an assistant can validate agent-tool policy files and audit signed tool-call traces.
Tokentrust
Published by GitHub account RudrenduPaul
A local MCP server that lets an assistant run TokenTrust's verification of an AI context-reduction proxy's claimed token and cost savings against a labeled task
Tenantguard
Published by GitHub account RudrenduPaul
TenantGuard exposes a scan tool that audits a self-hosted multi-tenant AI-agent deployment's configuration for tenant-isolation gaps.
Teamspend
Published by GitHub account RudrenduPaul
Teamspend is a local CLI that compares AI coding tool spend before and after a migration, pulling cost numbers from vendor admin APIs or local session logs.
Taskswarm
Published by GitHub account RudrenduPaul
Wraps the TaskSwarm CLI in a single generic tool so an assistant can coordinate and report status for parallel coding-agent sessions.
Slop Eval
Published by GitHub account RudrenduPaul
A local MCP server that wraps the slop-eval CLI, letting an assistant score AI-generated UI screenshots or URLs for genericness using an Anthropic LLM judge.
Skillguard
Published by GitHub account RudrenduPaul
A local MCP server that scans third-party AI agent-skill files (SKILL.md, hooks, scripts) for known attack patterns before an assistant installs or runs them.
Shimguard
Published by GitHub account RudrenduPaul
A local MCP server that lets an assistant run the ShimGuard CLI to check whether GitHub issues closed as fixed actually cite a merged pull request.
Ownvoice
Published by GitHub account RudrenduPaul
Wraps the OwnVoice CLI so an assistant can run its check, train and infer commands to build and use a LoRA voice adapter for pocket-tts.
Neuronscope
Published by GitHub account RudrenduPaul
Local MCP server that traces which attention heads and MLP neurons in a language model drove a predicted token, returning schema-versioned JSON.
Memtrust
Published by GitHub account RudrenduPaul
Runs memtrust's reproducible benchmark suite against agent-memory backends and returns the JSON report, so an assistant can compare memory systems on request.
Llmscout
Published by GitHub account RudrenduPaul
LLMScout is a local MCP server that runs 21 technical-SEO and generative-engine (GEO) checks against a live site and returns PASS/WARN/FAIL results with fixes.
Inferbench
Published by GitHub account RudrenduPaul
InferBench's MCP server lets an assistant run local LLM inference benchmarks on your own hardware and get measured tokens/second results back.
Graphkeeper
Published by GitHub account RudrenduPaul
A local MCP server that lets an AI coding agent build and query a graph of which files in a git repo historically change together.
Evolveguard
Published by GitHub account RudrenduPaul
A local MCP server that exposes the evolveguard CLI as one generic tool, letting an assistant record and check Claude Agent Skill files for capability drift.