Best Developer Tools MCP Servers
MCP servers for developer tools, what each one reports it can do, and whether Pod has verified it.
405 servers, ordered by adoption — npm downloads where a package exists, then reported issue volume. Pod has connected to 3 of them; the rest are publisher-reported and unverified.
Showing 201–225 of 405 · page 9 of 17.
Pensieve
Pensieve builds a context layer for your company.
It connects to the tools your work already lives in, including Slack, Notion, Google Drive, GitHub, HubSpot, Attio and Linear, and continuously curates what it finds into a navigable map of Pages: what your product does, who your customers are, what was decided and why.
This connector gives Claude access to that map. Instead of searching each tool separately and stitching the results together, Claude reads a layer that has already been curated across all of them. Every answer cites the source documents it came from, so you can check the working.
Claude can also write back. It can capture a decision from a conversation, correct a Page that has gone stale, or add structure as your company changes.
Pensieve keeps the layer current on its own. When a document changes in a connected tool, the affected Pages are revisited, so what Claude reads reflects the company as it is now rather than as it was when it was first connected.
16 tools reported — Publisher-reported; not verified by Pod.
Railway
The Railway connector lets Claude work with your Railway workspaces directly.
Claude can take actions: create new projects, deploy services from your connected GitHub repositories, redeploy services, manage feature flags, and publish self-contained artifacts (a static page, a React component, a diagram) to a public URL in seconds via Railway Functions.
It can also list your projects, services, and deployments, check the live status of an environment, and pull build, deploy, and HTTP logs to help you debug production issues without leaving the conversation.
For complex, multi-step operations, the connector includes railway-agent, a conversational bridge to Railway's own infrastructure agent, which can inspect services, diagnose failures, and apply configuration changes on your behalf. Destructive actions are gated behind explicit confirmation, and everything is scoped to the projects and workspaces your Railway account can access.
16 tools reported — Publisher-reported; not verified by Pod.
AccessLint
AccessLint monitors and scans websites for accessibility problems.
Most accessibility problems live behind an interaction. AccessLint saves user journeys that cross multiple pages and interactions, that a real browser replays and checks on a schedule. AccessLint can also scan a public page ad hoc against WCAG. In either case, it returns rich findings that you can dig into with Claude and file issues or spin up remediations.
What you can do:
- Scan a public page against WCAG and get the violations back.
- Turn plain sentences about user journeys, for example "search for a product and add it to the basket", into flows, with the steps compiled and auto-healed from the sentence.
- Look at the screenshot of the page as the browser saw it, at any step.
- Review the history and open accessibility violations a flow is currently finding
Requires a free AccessLint account. Single page scans work out of the box; you must verify ownership of a domain to run flows against it.
15 tools reported — Publisher-reported; not verified by Pod.
api.agi.tech
An MCP that can use your entire phone. It can use every app, without any integrations or connectors. It gives your agent access to tap and type on your screen, just like you do. No integrations to build, and nothing to configure per app. Android now live, iOS coming soon.
Use cases:
- Connect your phone to your agent OpenClaw, Hermes, Eve, Cursor, Claude Code, etc
- Developer testing: test your Android app under real-world conditions on a variety of screen sizes
- Automations: do recurring task on phones
15 tools reported — Publisher-reported; not verified by Pod.
ESO
Elite Site Optimizer helps businesses analyze and improve their websites through SEO, accessibility, performance, security, content, and AI-readiness audits, with actionable recommendations for better visibility and user experience.
15 tools reported — Publisher-reported; not verified by Pod.
Hiveflow
Hiveflow is a visual AI agent orchestrator: you design workflows on a drag-and-drop canvas, connect them to your tools and data, and let them run on a schedule, from a trigger, or on demand. This connector brings your Hiveflow workspace into Claude, so you can inspect and operate it in conversation instead of switching to the dashboard.
What you can do with it:
• Flows — list your workflows with their execution state, inspect a flow's nodes and edges, create new flows, run one on demand, and activate or pause it. • Kanban boards — list boards, read columns and cards, create cards, move them between columns, and add comments. • Workers — see which agents are connected to a board, whether they are paused, and when they last checked in. Moving a card into a watched column is how you hand work to them. • Organizations and workspaces — list the organizations and workspaces you belong to, and scope any operation to a specific one. • Live API reference — a docs tool that returns the current Hiveflow API reference, so Claude works from up-to-date definitions rather than assumptions.
Everything runs with your own Hiveflow account through OAuth: the connector sees exactly what you see, respects your organization roles and permissions, and every action is attributed to you. Tools that only read are marked read-only; tools that change your account — running a flow, pausing one, moving a card — are marked as such, so Claude asks before acting.
Free to connect with any Hiveflow account.
15 tools reported — Publisher-reported; not verified by Pod.
Matia MCP
Matia is the Unified DataOps Platform for data, AI, and engineering teams, combining ingestion, reverse ETL, observability, and catalog in one place. This MCP gives your agent a single interface to all of it.
Because Matia is one unified platform rather than a stack of point tools, the MCP can follow a problem across the entire data lifecycle in a single session. It answers where a number came from and what breaks if you change it, without stitching together separate tools for ingestion, observability, and activation.
Works alongside your warehouse MCP, such as Snowflake, so agents move from pipeline health to the underlying data in one flow. Requires a Matia account.
Better together. Do more with Matia.
15 tools reported — Publisher-reported; not verified by Pod.
Mintlify
Connect Mintlify to Claude to keep your documentation accurate and up to date without leaving your conversations. Claude can search across every page, read and edit content directly, and ship changes. Perfect for fixing outdated pages after a product change, reorganizing sections as your product grows, or drafting new content alongside the code it documents.
15 tools reported — Publisher-reported; not verified by Pod.
Mnemom
Mnemom is trust infrastructure built for agents, not just the humans who run them. In one chat you can look up an agent's Continuous Trust Rating — the score, the grade, and the accountability evidence behind it — and confirm it's cryptographically signed, not merely asserted. Drop your own alignment card and get an instant trust read back; claim a verifiable identity with no human in the loop; and walk away with an embeddable Trust Rating badge for your README or agent card. Every read is honest about what it knows: a quiet record reads as quiet, never as a red flag. The system tells the truth — and view-source proves it.
15 tools reported — Publisher-reported; not verified by Pod.
Montra
Montra turns narrated screen recordings into structured tickets with annotated screenshots and transcripts; this connector lets Claude browse your workspace's documents and tickets, pull the screenshots and transcripts behind each one, and create, update, link, and close tickets as it implements the work.
15 tools reported — Publisher-reported; not verified by Pod.
Outtake
Outtake finds and removes threats to your brand: impersonation accounts, phishing domains, counterfeit listings, and executive impersonation across social platforms, app stores, domains, and the open web.
Connect this server to bring your Outtake workspace into Claude. Ask what changed this week and get the digest your analysts read, with the alerts behind it. Pull the full analysis payload for a digest when you want to summarize it, chart it, or send it somewhere else. List the entities you protect and check where they're being abused.
You can also act. Submit a suspicious URL for review and enforcement, preview a takedown before filing it, file it, then track delivery and platform confirmations without leaving the conversation.
Access is scoped to a single Outtake profile that you choose when connecting, and to your own permissions on that profile. Requires an Outtake account.
15 tools reported — Publisher-reported; not verified by Pod.
PlanetScale
Access to your PlanetScale organizations, databases, branches, schema, and Insights data.
15 tools reported — Publisher-reported; not verified by Pod.
Subtext
Subtext is session replay built for agents instead of humans. Connect it and your Claude agents can open real production sessions and investigate them directly—no scrubbing a timeline, no video to watch. Learn more at https://subtext.fullstory.com/docs/introduction
Hand your agent a Subtext link from wherever it already looks—a Sentry issue, a Linear ticket, a support thread, a PR—and it jumps to the right moment, inspects screenshots and accessibility trees, diffs DOM state, and reads console and network activity to find the root cause. Then it files the fix or hands you the finding. You review the answer, not the replay.
What your agent gets:
- Session tools it calls directly: open a session, jump to a moment, inspect and compare state—purpose-built so agents traverse dense session data efficiently, with fewer tokens than pointing a general agent at raw replay.
- Best-in-class capture from Fullstory: Fullcapture behavior, pixel-perfect screenshots on demand, DOM diffs, and windowed console and network context.
- Privacy built for production: precision tools for agents to mask or exclude sensitive fields, plus agent-friendly DLP scanning tools over MCP.
- Cost-effective: Capture every session without sampling and never miss a critical user experience or bug report. Agents review the ones that matter most.
Works with Claude Code or any MCP-compatible agent. Code-first install, self-serve signup, generous free tier, usage-based pricing. Bring your own agent harness—Subtext is the session context it was missing.
Session replay used to be something humans watched. Now it's evidence your agents act on.
15 tools reported — Publisher-reported; not verified by Pod.
Tempreon
Tempreon is a personal AI operating system: a person-owned intelligence layer that captures how you think, decide, write, and work — your identity, knowledge, preferences, decision rules, guardrails, and accumulated lessons.
It's more than memory. Memory recalls what you said; Tempreon learns what you meant and how you decide. Every session generates observations, patterns are tested against real outcomes, and what holds up gets promoted into validated instincts that shape future work. Your AI doesn't just remember more — it gets better at being yours.
Connected to Claude, every conversation starts with continuity and judgment. Claude knows who you are and what you're working on, searches your personal knowledge and File Vault, manages your tasks, and can dispatch your trained agents — named specialists that carry your voice — for autonomous, audit-trailed work. And the drafts, outreach, and replies you write together hold your voice — cadence, vocabulary, non-negotiables — so the result reads like you, not a template.
Because the layer is person-owned, it's portable: the same identity, learning, and guardrails travel with you across engines and tools. Your context isn't locked inside any one assistant. It's yours, applied wherever you work.
Scoped to you alone: your data is isolated to your account, never sold or shared, and writes happen only on your explicit action.
15 tools reported — Publisher-reported; not verified by Pod.
Credits.fm
Credits.fm is a public MCP server that lets AI agents look up and resolve music credits and identifiers in real time. It supports fast searches and normalization across core music IDs including ISRC, ISWC, IPI, ISNI, and UPC, helping agents match recordings to works and rightsholders with higher confidence. Use it to enrich metadata, validate identifiers, and power music-aware workflows without authentication. Designed for lightweight, high-signal lookups that can be embedded directly into agent tools and automations.
14 tools reported — Publisher-reported; not verified by Pod.
iknow.dev
iknowは、AIアシスタントとの協働を前提としたナレッジ管理プラットフォームです。構造化されたナレッジを「エージェント」単位で保持し、MCPサーバー経由でLLMがプログラム的に読み書きできます。 主な機能:
キーワード検索・セマンティック検索・ハイブリッド検索によるナレッジ探索 Markdownとテーブル形式(JSON)の両方に対応した柔軟なナレッジ登録 タグ付けとナレッジ間のリンクによる構造化・整理 1アカウントで複数の独立したエージェント空間を管理可能 楽観的ロック(expected_version)による書き込み競合の防止
活用例:
AIアシスタントからアクセスできる個人・チームのナレッジベース Claudeなど各種LLMから問い合わせ可能なチームWiki 予算表・機能リスト・チェックリストなど構造化データのAIアクセス プロジェクトの進行に合わせてAIが更新できるドキュメント
iknow-mcpはJSON-RPC over stdio/HTTPに対応した標準的なMCPサーバーで、ツール定義・入力スキーマ・エラー応答についてMCP仕様に準拠しています。
14 tools reported — Publisher-reported; not verified by Pod.
Lightrun
Lightrun MCP connects your AI assistant to real runtime data from applications already running in production, staging, or development. Built on the Model Context Protocol, it lets Claude investigate live code behavior on your behalf—without changing code, rebuilding, or redeploying.
What you can do
- Discover connected agents, agent pools, and tags in your environment
- Capture expression values and call stacks at specific code locations (snapshots)
- Count how often a line executes
- Measure execution duration between two lines (TicToc)
- Sample custom numeric metrics from live code
- Drill into raw measurement samples for deeper analysis
Why use it Lightrun brings production-safe, on-demand observability into natural-language workflows. Instead of configuring probes manually, describe what you need and let the AI discover targets, place temporary instrumentation, and analyze results. Runtime data is subject to your organization’s PII redaction and permission rules.
How it works
Your AI assistant acts as an MCP client and connects to your Lightrun server at /mcp. The Lightrun MCP server translates requests into safe runtime inspections executed by Lightrun agents attached to your applications.
Authentication
- OAuth for interactive use (sign in with your Lightrun user)
- API key (Bearer token with Dev scope) for agents and automation
Works great with Lightrun AI Skills Pair this connector with reusable investigation workflows from the Lightrun AI repository—preflight checks, evidence collection, diagnosis, and handoff output for live runtime debugging.
Requirements
- A Lightrun account and server (SaaS: https://app.lightrun.com)
- MCP access enabled for your organization
- Lightrun agents connected to the applications you want to inspect
14 tools reported — Publisher-reported; not verified by Pod.
SimpleBackups
SimpleBackups is automated backup management for databases, servers, cloud storage, and SaaS apps like Notion, GitHub, Slack, and Linear. This connector links your SimpleBackups account to Claude so you can check backup health, investigate failures, and run safe operations without opening the dashboard.
What you can do:
- Get an account-wide health check in one call: how many backups ran, which failed, which have gone stale.
- Diagnose a specific failure using the actual console output from the run, not a truncated summary, so you get the real cause (bad credentials, disk full, permission denied, timeout).
- Chart run history across all your backups over the last 7 to 30 days.
- Trigger an on-demand run, pause or resume a schedule, and prepare a restore that returns a short-lived signed download link or an rsync command you run yourself.
Example prompts:
- "Did any of my backups fail in the last 7 days?"
- "Why did the production Postgres backup fail last night?"
- "Show me the run history for all my database backups this week."
- "Run the staging MySQL backup now."
- "Prepare a restore of yesterday's Notion backup."
How it is built: 10 read-only tools and 4 write tools. Every write tool is reversible and requires a Member role or higher on the connected team. No tool ever returns credentials, API tokens, or sensitive data. Restore links are time-limited and expire on their own; SimpleBackups never runs a restore on your behalf from chat. Access is per-team, chosen during sign-in, and revocable at any time from your SimpleBackups settings.
Requires a SimpleBackups account on a trial or paid plan.
14 tools reported — Publisher-reported; not verified by Pod.
Vertiso Memory
Vertiso Memory is a portable, user-owned memory layer for AI agents. Connect it once, and your assistant can capture and recall durable facts, decisions, preferences, and project context across every tool you use, rather than starting from scratch in each new conversation.
It exposes a small, composable toolset over MCP: remember to capture a durable fact, recall for hybrid search with synthesized answers, search for a ranked hybrid record lookup, and get_memory to inspect specific memories. Session tools like hello, handoff, and wrap-up let work continue cleanly across sessions and devices.
Your data is yours. We do not train AI models on your content. Memories are sent to third-party LLM providers solely for retrieval and synthesis, under those providers' standard no-training API terms, and are deletable upon request.
Works with any OAuth-capable MCP client. Connect by adding the server URL; client registration and sign-in happen in the browser, with no API key to paste.
Available worldwide except where prohibited by applicable law or unsupported by our service providers.
14 tools reported — Publisher-reported; not verified by Pod.
Windmill
Search for scripts and flows, get their details, and execute them directly from Claude. Monitor job executions, retrieve logs, manage resources and variables, and control schedules. Windmill is an open-source platform for internal tools, workflows, and API integrations, supporting Python, TypeScript, Go, SQL, and more. Works with Windmill Cloud or self-hosted instances.
14 tools reported — Publisher-reported; not verified by Pod.
Arkon Vault
Arkon Vault is a private AI work-continuity layer for people and teams using more than one AI. Claude can preflight work, find the authoritative project and next action, accept or create structured handoffs, publish checkpoints, preserve decisions, search shared context, and save project state. The MCP connector stores continuity content a user deliberately sends through Vault tools; it does not automatically ingest the surrounding Claude conversation or read local files. Optional Arkon Sync is separate, opt-in Windows software. It sends limited session-derived project records from supported local AI sessions—including title, a brief summary derived from session metadata or the first user message, AI platform, last-active time, message count, working directory, deep link, and session ID—to Arkon Vault's servers, where they are stored in the user's tenant-isolated private vault. It does not upload complete raw transcripts or full assistant responses. Users can view records in the Vault dashboard, download a tenant-scoped JSON export from Account → Your data, or submit or cancel a verified deletion request there. Details: https://vault.arkoncybersecurity.com/data-deletion. The MCP vault_sync tool only refreshes the server-side continuity index and any configured Notion mirror; it cannot start Arkon Sync or upload AI conversations.
13 tools reported — Publisher-reported; not verified by Pod.
Autorender
Autorender is image and video infrastructure built for AI app builders — reliable uploads, fast CDN delivery, and storage/optimization handled for you. This connector brings your real Autorender library into your workflow: browse and search files by folder or tag, upload new images and video (including large files via multipart upload), organize folders, and update metadata — all against your actual account data, not mock placeholders. Every call runs through your own Autorender API key via OAuth, so builds use real CDN URLs from the start instead of stand-in images.
13 tools reported — Publisher-reported; not verified by Pod.
Decionis Execution Authority
Decionis is an execution-authority layer for AI agents. Before an agent takes a consequential action, Decionis evaluates the proposed action against organization policy and determines whether it may proceed, must be held, or requires additional authority.
When human approval is required, Decionis can request verified approval through Presence and re-evaluate the action using the resulting signed evidence. Every evaluation produces a verifiable Decision Dossier and cryptographic evidence for audit and accountability.
Use Decionis to govern agent-initiated payments, refunds, commerce operations, enterprise workflows, infrastructure changes, and other high-impact actions.
Decionis authorizes execution; it does not perform the downstream action. Payments, transfers, trades, orders, and system changes remain the responsibility of the connected executor.
13 tools reported — Publisher-reported; not verified by Pod.
icod.ai
From prompt to live production. Turn Claude into your development agent by sending optimised instructions straight to icōd.ai to build and update apps instantly.
13 tools reported — Publisher-reported; not verified by Pod.
InstaPods
Getting a small app online normally means renting a server, installing a runtime, configuring nginx, setting up SSL, a firewall and a process manager, then pointing a domain at it. InstaPods skips all of that. Pick a runtime, get a real Linux server with a live HTTPS URL in seconds, from $3 a month flat with no surprise bills.
This connector puts that in the chat window. You can see every pod on your account and whether it is running, deploy a new one from a preset (static, PHP, Node.js or Python), read the logs and files when a site starts failing, fix the file and reload the app, run a shell command inside your own pod, and move a pod up or down a plan when it needs more or less CPU, memory and disk. So "why is my site returning 502" and "put it back up" become the same conversation instead of a detour into a terminal.
Thirteen tools: eight read-only ones for listing pods, presets, plans and regions and for reading files and logs, and five write ones for creating a pod, managing its lifecycle, changing its plan, writing a file and running a command. Read-only tools are annotated as such so they run without a confirmation prompt; every write tool that modifies or removes something prompts you first.
Every action is scoped to the team that owns the OAuth session, so you only ever see and change your own pods. Plan selection and payment happen on instapods.com, never in chat.
13 tools reported — Publisher-reported; not verified by Pod.
Page 9 of 17. ← previous · next →
For agents
This page has a Markdown and a JSON twin. Pod is also queryable over MCP at https://api.askpod.ai/mcp/read.