MCP server directory

Gives an agent scoped read, write, and search access to directories you explicitly allow. It is the reference implementation for local file work: you pass one or mor…

Read-only access to a repository's history: status, diffs, commit logs, and blame. Useful when an agent needs to understand why a file looks the way it does before c…

A persistent knowledge graph stored in a local JSON file, exposed through entities, relations, and observations. It is the simplest way to give an agent recall acros…

Downloads a URL and converts HTML into Markdown so the model reads text instead of markup. Handles chunking and start-index paging for long documents. It respects ro…

Timezone conversion and current-time lookup using the IANA database. Small, boring, and quietly essential: models routinely guess the wrong date, and a single correc…

Gives the model a scratchpad for structured, revisable reasoning: numbered thoughts that can branch, revise, and conclude. It does not make a weak model strong, but …

A deliberately exhaustive reference server that exercises prompts, resources, tools, sampling, and notifications. You do not install it to get work done; you install…

Official GitHub server covering repositories, issues, pull requests, code search, and Actions. Ships as a remote endpoint or a local binary, and lets you scope an ag…

Drives a real browser through accessibility snapshots rather than screenshots, so an agent can click, type, and assert without a vision model. This is the practical …

Screenshot, navigate, and run JavaScript in a headless Chrome. Archived, and worth knowing about mainly because a large number of tutorials still point at it. For ne…

Runs read-only SQL against a Postgres database and returns the schema alongside results, which is what makes it usable without pasting table definitions into the pro…

Query and inspect a local SQLite file, including schema discovery and arbitrary SELECT statements. The lowest-friction database server for experiments, and a good fi…

Reads channel history, threads, and replies, and can post messages on your behalf. Archived upstream; most teams now reach for a vendor-maintained Slack server or ca…

Adds web and local search through the Brave Search API, returning titles, snippets, and URLs the agent can then fetch. Requires your own API key. Archived, but still…

Geocoding, place search, distance matrices, and directions via the Google Maps API. Needs a billing-enabled key, so treat it as an integration rather than a weekend …

Pulls issues, stack traces, and event context into the agent so it can triage production errors with the actual payload in front of it. Remote-hosted, which means no…

Serves version-specific documentation for thousands of libraries, so an agent stops writing code against an API that changed two releases ago. The practical fix for …

Scrapes, crawls, and maps sites, returning clean Markdown with JavaScript already executed. Handles the pages a plain fetch cannot: infinite scroll, client-rendered …

Manages Supabase projects from an agent: run SQL, inspect tables, apply migrations, and manage edge functions. Powerful enough to be dangerous, so use a scoped perso…

Neural web search that returns full page contents alongside results, which removes a second fetch step for research agents. Better than keyword search for conceptual…

Runs Apify actors on demand and hands their output back as context: scrapers, crawlers, search, and document extractors behind one interface. The hosted endpoint han…

Drives real iOS simulators and Android devices: tap, type, scroll, and read the accessibility tree rather than pixel coordinates. The practical way to run a mobile f…

Terminal access, process control, and file editing in one server, including a diff-based editing mode that avoids rewriting whole files. It is the closest thing to h…

Connects IDA Pro to an agent so it can query functions, cross-references, strings, and decompiled output while you analyze a binary. Requires IDA running locally wit…

A security-focused server that exposes a large catalog of offensive tooling to an agent, from reconnaissance through exploitation support. Self-hosted Python service…

Gives an agent the n8n node schema — hundreds of nodes with their real fields — so it can build and debug workflows instead of inventing parameters. The documentatio…

Lets an agent create and edit draw.io diagrams instead of describing them in chat. The server works on the diagram model rather than the rendered image, so moving a …

Sandboxes large tool outputs before they reach the model and keeps a persistent session memory, aiming to reduce context consumption in long agent runs. The vendor c…

Vector memory with no server to keep running: a single .rvf file holds the embeddings and the CLI exposes it over MCP. The appeal is portability — commit the file, c…

Covers the GitLab surface an agent needs day to day: merge requests, issues, pipelines, releases, and wiki pages. Authenticated with a personal access token, so scop…

Drives a live Chrome instance over the DevTools protocol: performance traces, network requests, console output, and accessibility snapshots. It answers the question …

Exposes collections, indexes, and aggregation pipelines so an agent can inspect a database instead of guessing its shape. The setup command writes credentials to a l…

Key inspection and read queries against a Redis instance: scan by pattern, check TTLs, and look at what is actually stored. The value is diagnostic — comparing the c…

Searches indices and reads mappings through the Elasticsearch API, so an agent can debug a query against the real cluster rather than a remembered schema. Ships as a…

Reads Terraform provider documentation and registry metadata so an agent writes HCL against the real schema instead of inventing arguments. It does not run plan or a…

Reads and writes pages and databases in a Notion workspace through the official API. Requires an integration token and, just as importantly, a page shared with that …

Pulls file structure, components, and design tokens out of Figma so generated code matches the design instead of approximating it. The value is in the tokens — color…

Talks to the Kubernetes API directly instead of shelling out to kubectl, so an agent can list contexts, read pod logs, and apply manifests without parsing command ou…

Exposes Azure DevOps to an agent: work items, repositories, pull requests, and pipelines. The practical use is triage — reading a failing build and the work item it …

Connects an agent to a running Next.js dev server so it can read routes, build output, and runtime errors instead of guessing from source alone. Useful during a refa…

Drives xcodebuild and the simulator from an agent: build, test, launch, and read logs. This is the missing piece for iOS work, where the build loop otherwise means a…

Gives an agent the dependency graph of an Nx monorepo, so it can answer which projects a change actually affects. On a large workspace that question is the differenc…

Brings SAP Fiori application scaffolding and configuration into an agent, including the generator options that are otherwise buried in documentation. Narrow by desig…

Wraps the Azure CLI surface — resource groups, storage, Cosmos DB, Key Vault — so an agent can inspect a subscription and propose commands with the real names in the…

Exposes gcloud operations to an agent: listing projects, inspecting resources, and running read commands against Google Cloud APIs. Best used against a sandbox proje…

Lists buckets and objects and pulls metadata for a given path. Useful when the question is 'what is actually in this bucket' rather than 'download it'. Content reads…

Queries Cloud Logging and Monitoring so an agent can answer 'what changed at 14:02' with actual log lines instead of a hypothesis. On an incident this is the differe…

Bridges Microsoft Graph: mail, calendar, files, and the directory. Requires an Azure app registration and admin consent, which is the real work — the server itself i…

Serves SAPUI5 and OpenUI5 API documentation plus project scaffolding to an agent, so generated controls use real properties instead of plausible ones. Narrow, and th…

Reads variables, styles, and component trees from Figma and can write changes back. The token extraction is the useful half; the write half needs a scoped token and …

Exposes Kubb's code generation as a tool, so an agent can regenerate a client from a spec instead of hand-editing forty files. Valuable exactly when the schema chang…

Reads and changes feature flags through the LaunchDarkly API, so an agent can turn a flag off during an incident instead of asking a human to open a dashboard. The r…

Searches messages, threads, and channel metadata in a Slack workspace, and can post when given a token with write scope. The search half is immediately useful; the p…

Queries the Datadog API for metrics, monitors, dashboards, and incident data. On an incident this turns 'something is slow' into an actual series with a timestamp. C…

Talks to Metabase over its API with a key rather than OAuth, which makes it easy to wire up and easy to over-permit. Useful when the question is 'what does this save…

Serves current Mastra documentation, changelogs, and examples to the agent, which is the difference between working code and plausible code when a framework ships we…

Brings CAP documentation and project conventions into the agent so generated entities and service definitions follow the framework instead of resembling Express. Lik…

Introspects a GraphQL endpoint and lets the agent compose queries against the real schema, including variable handling and per-request headers. That last part matter…

Stores and retrieves product knowledge — decisions, specs, context — so an agent can answer 'why is this built this way' without reading six months of tickets. The v…

Reads and writes Linear issues, projects, and cycles through the GraphQL API. The useful part is not creating an issue — it is filing one with the right team, label,…

Exposes Salesforce orgs to an agent: describe objects, query records, and inspect metadata. The hard part is never the connection, it is the permission model — an ag…

Official server for Xero accounting data: invoices, contacts, payments, and bank transactions. Finance is the category where a wrong write costs more than a wrong re…

Web search built for agents rather than people: results come back with extracted content and optional answer synthesis, which removes a second fetch step. Requires a…

Reads Stripe objects and can create test-mode resources, which makes it genuinely useful while wiring a checkout. The warning is the same one Stripe gives: use a res…

Queries the Clarity data export API for session metrics, heatmap data, and page behavior. Useful when the question is 'where do people stop reading' and the answer s…

Official stdio server for Discourse forums: search topics, read posts, and inspect categories. The obvious use is answering support questions from past threads. Keep…

Manages Unleash feature toggles: list flags, read strategies, and flip state. The read path is the safe one and covers most of what an agent needs during a release. …

Brave's own server, and the one to use now that the older reference implementation sits in the archived repository. Returns web, news, image, and local results as st…

Attaches to a real Chrome profile instead of launching a headless one, so the agent sees the pages you are logged into. That is the whole value and the whole risk: i…

Runs Deque's axe engine against a page and returns structured violations with the failing selector and rule. It will not find everything — automated checks catch rou…