PurseString
Local installSpend policy and human approvals for AI agents, checked before a single-use capped card is issued.
Every server in the official MCP Registry. Ranked by whether they actually work: we connect to each one and publish what happens.
61,546 servers
Spend policy and human approvals for AI agents, checked before a single-use capped card is issued.
Use your OTPilot accounts (passwords, 2FA codes) to sign in and run commands, without seeing them.
Validate, parse, generate and convert EU e-invoices (XRechnung, ZUGFeRD, Factur-X, Peppol BIS).
Read and send texts over iMessage, RCS and SMS from your Blujay phone numbers.
Google Workspace automation for AI agents: 39 tools, human-approved deletes, bulk undo.
iMessage and SMS for AI assistants, from a Mac you control. Read, search, send with approval.
Windows terminal sessions for agents: spawn, type, wait without polling, lean reads, screenshots.
Research X (Twitter), read inboxes and review publishing through MCP. No X API key required.
Find Python dependencies with no prebuilt wheel for your platforms: ARM, Alpine, Windows on ARM
Self-hosted remote MCP server for Docmost CE: members read and edit the wiki as themselves
Deploy self-hosted OpenClaw to AWS, GCP, Azure or a local VM — a CLI and an MCP server in one.
PassProfessor exam prep: program catalog, verified exam facts, and free sample questions.
Policy checks for AI agent writes (ALLOW/DENY/STEP_UP) with a tamper-evident audit log.
Search, resume and hand off coding-agent sessions: Claude Code, Codex, Gemini, Antigravity and more.
Checks an answer against its sources: numbers and citations verified, wrong ones fixed. No model.
Send big files and folders with a link, manage transfers and file requests on pandafile.co.
Czech company registry: search, profiles, relationships, status and evidence by ICO.
MCP server for SSH operations with built-in diagnostics
Turns rough prompts into detailed, LLM-ready prompts with multiple-choice clarifying questions.
Search a directory of software an agent can adapt or operate, by need, license or interface.
Publish drill-down diagrams from an agent; every Mermaid subgraph becomes a level you enter.
Query Playwright test analytics from AI assistants: flaky tests, failures, CI time lost to retries.
Play Sid Meier's Civilization V (native Linux) as a human seat, by the game's own rules.
Diagnose a repo for AI agents: context noise, missing commands, cache value, flaky tests.
Point-in-time exchange trading rules with sources: A-share price limits, stamp duty, ticks, futures
Auditable agent memory: lessons cite their sources, and bad memories can be traced and removed.
Bookmark memory for AI agents: save links, search inside saved pages, recall what you read.
Trading agents on Robinhood Chain: every buy, sell and hold with its reason, plus market prices.