Grounded Reasoning
io.github.ALEXaquarius/grounded-reasoning
Verify multi-hop relational claims before an agent asserts them — zero tokens, with proof paths.
OLACHEAP · MCP WORKSPACE / DEVELOPER
Choose developer MCP servers for one concrete job: reading repositories, tracing pull requests, reviewing code changes, or checking build status. A tool that can search code is not automatically appropriate for merging a pull request.
Results come from the current synchronized Registry snapshot, not popularity or activity.
io.github.ALEXaquarius/grounded-reasoning
Verify multi-hop relational claims before an agent asserts them — zero tokens, with proof paths.
io.github.AM-Digital-Research-Environment/amira-mcp-server
Read-only access to AMIRA, the Africa Multiple research-data platform on Omeka S.
https://data.africamultiple.uni-bayreuth.de/mcpio.github.AMMAR-62/mewcp-google-calendar
Schedule smarter - create, search, and manage Google Calendar events through AI.
io.github.AMRITO-KUNDU/llmscribe
Local project map, search and file reader for AI agents.
io.github.ANAMIZED/lrsi
Local recursive self-improvement OS MCP with fail-closed governance and gated mutation tools.
io.github.ANAMIZED/needrail
Needs registry for public goods (MCP + x402): discover, fund, claim, complete with provenance.
io.github.ANAMIZED/opengos
Grants and public-goods MCP: discovery, ranking, grounded drafting, open-source first.
io.github.ANAMIZED/openmesha
Open agentic operations mesh MCP: agents, workflows, cost, governance, and x402.
io.github.ANAMIZED/server-os
Fail-closed agentic OS MCP: agents as processes with cost, governance, observability, skills.
io.github.ANAMIZED/superagenticmcp
MCP-native router and multi-agent control plane: plan, route tool calls, stream execution.
io.github.ANAMIZED/yodmcp
Autonomous MCP Agent OS: multi-graph memory, Tasks, Skills, A2A, attestation, OpenTelemetry.
io.github.ANHQ280/myinvois-mcp
Malaysian LHDN e-invoicing (MyInvois) compliance rules: scope, validation, change-diffs.
io.github.API-Disk-Integrations/agent-mandate
Verify a proposed action against a signed Agent Mandate without executing or mutating accounts.
io.github.APIHubsRU/apihubs
Generate AI video, images and music. Browse models and prices; OAuth and one prepaid balance.
https://api.apihubs.ru/mcpio.github.APIStock/api-stock
Generate AI video, images and music. Browse models and prices; OAuth and one prepaid balance.
https://api.api-stock.com/mcpio.github.APMC1/xaps
Pre-execution audit for autonomous agents. Independent receipt before irreversible actions.
io.github.APProj/vendor-status-mcp
Status-page map for 1,100+ SaaS/cloud vendors plus 15,700+ incidents. No API key.
io.github.APProj/warn-mcp
Search 61,428+ US WARN Act layoff notices: 48 states, 1988-today, daily. Bundle needs no API key.
https://mqpacuiiqcnwk93jy.apify.actor/mcpio.github.APonkratov/ya-yandex-wiki-mcp
Yet Another Yandex Wiki MCP Server.
io.github.APonkratov/yandex-tracker-mcp
MCP server for Yandex Tracker API.
io.github.ARAS-Workspace/claude-kvm
MCP server — control remote desktops via VNC with a native Swift daemon and Apple Vision OCR
io.github.ARCXS-Protocol/mcp-server
The DNS/SMTP for AI agents: cross-protocol discovery, messaging and transactions. No middleman fees.
io.github.AS-XS/ai-pensieve-mcp
Search and cross-reference local AI chat history with read-only tools and source citations.
io.github.ASCIT31/darkmoon-mcp-server
Drive a self-hosted Darkmoon Pro AI pentest instance: start runs, read campaigns and findings.
Synchronized data; no invented downloads, reviews or ratings.
OLACHEAP · PRACTICAL GUIDES
Start with a read-only repository and ask the AI client to locate a known file or explain an existing change. Confirm the answer against the actual commit and file path before enabling changes to branches, issues, or CI settings.
Compare the source repository, authentication method, documented scopes, tool descriptions, and whether operations write to a repository. Prefer explicit approval of each merge, push, or deletion, and check which GitHub identity will own the change.
A successful repository search does not prove the server is secure. Treat issue descriptions, code comments, and README content as untrusted input, avoid exposing secrets in logs, and keep the ability to revoke tokens.
Use their exact Registry IDs to compare published connection information before granting permissions.