A-Modular-Kingdom
Allows AI agents to perform web searches using DuckDuckGo, providing current events and latest documentation.
Provides vision analysis and text embeddings using local Ollama models, enabling private and offline AI capabilities.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@A-Modular-Kingdomsearch my codebase for authentication logic using v2 RAG"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Elpis
Elpis is a Linux coding-agent environment built on OpenAI's Codex CLI. It keeps the execution and terminal foundation while adding control over admitted context, durable memory, continuity and provider selection.
Change the runtime. Keep the thread.
Releases · Website · Technical guide · License
Install
v0.4.1 is the Linux x86_64 release, based on Codex rust-v0.160.0. Review the installer before running it:
curl -fsSL https://raw.githubusercontent.com/MasihMoafi/Elpis/v0.4.1/scripts/install-elpis.sh | bash
~/.local/bin/elpisThe installer verifies SHA-256 checksums and installs Elpis, its Code Mode host and the bundled Linux sandbox. A Debian package is also available in the release. RTK is optional and is not installed automatically. macOS and Windows binaries are not included.
Sign in or configure a provider, then choose a model with /model.
Related MCP server: MCP Agent Memory
State and upgrades
Elpis uses ~/.elpis-next, or the directory explicitly selected by ELPIS_HOME. It ignores inherited CODEX_HOME. Existing next-build users keep their state there.
v0.3.0's state database is incompatible with the new foundation. Elpis refuses that database and preserves it; automatic migration is not included. Keep a backup of the old executable for rollback. The v0.3.0 guide records that release's behavior and historical evidence.
Included
Context Ledger controls admission of workspace instructions, chosen files and memory;
/addand/contextexpose the working set.The responding agent explicitly saves durable knowledge and the workspace checkpoint through a guarded local tool. Saving and admission are separate choices.
/compactaccepts a pressure threshold or custom instructions. Experimental Smart Prune and reasoning expiry limit disposable context./modeland--providerselect providers; the local gateway supports Anthropic, Gemini and OpenRouter paths alongside OpenAI./dashboardand/usageexpose local session information. Subscription price remains unavailable.Elpis identity, composer queuing, chosen skills and permission controls surround the selected model.
No model weights, retrieval engine or speech engine are bundled. External capabilities come through tools or explicitly registered integrations.
Verification and limits
Masih tested the September 30 candidate and accepted it for release. Release CI gates publication on focused Elpis Rust checks, actual runtime evals with failing controls, installer checks and clean offline binary/Debian installations. Automated evidence does not establish general coding-quality gains.
The whole inherited Codex suite is not claimed green: appearance/snapshot differences and identity-rename failures remain. /force-prune, Auto routing and tui.appearance are not ported. Strict configuration currently misses unknown keys in gateway-provider tables. Light-theme acceptance and all historical outcome-ledger cases remain separate from this release's acceptance.
Development
Read AGENTS.md, the guide, shipping rules and local build rules before changing or building Elpis. The upstream Rust workspace version stays at 0.160.0; the Elpis product version lives in the CLI manifest and TUI branding constant.
Upstream Codex is Apache-2.0. Elpis retains its attribution and notices. See LICENSE.
This server cannot be deployed
Maintenance
Related MCP Connectors
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
MCP server for building and testing AI agents with multi-model experimentation and insights.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
- UnifAPIOAuthcom.unifapi
Hosted MCP server for live public-data APIs and Skills for AI agents.
Related MCP Servers
- -licenseNot gradedqualityNot gradedmaintenanceA sophisticated MCP server providing advanced memory capabilities with RAG, hallucination detection, and enterprise-grade AI infrastructure for intelligent agent ecosystems.-
- AlicenseNot gradedqualityCmaintenanceA production-ready MCP server that enables multiple AI agents to collaborate through a shared, concurrency-safe memory space. It supports advanced search, full CRUD operations, and automatic backups to facilitate asynchronous communication between agents.MIT
- AlicenseCqualityDmaintenanceA production-grade MCP server providing an autonomous AI agent swarm, persistent semantic memory, browser automation, multi-model reasoning, and 78+ tools for AI-first testing and development.63MIT
- FlicenseNot gradedqualityDmaintenanceA robust, lightweight Model Context Protocol (MCP) server designed to empower your AI Agents with context-awareness, safe execution sandboxes, and dedicated thought logs.-