Log inSubmit server
Ejentum MCP logo

Ejentum MCP

by ejentum

Retrieves structured reasoning procedures for analytical, coding, integrity, and conversation-perception tasks through MCP.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Ejentum MCP

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "ejentum-ejentum-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "ejentum-mcp"
      ],
      "env": {
        "EJENTUM_API_KEY": "YOUR_VALUE_HERE"
      }
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Values are written into the snippet above so you can paste it as is. They never leave your browser.

Automated check passed— started and listed 8 tools correctly (1mo ago).

Overview

The ejentum/ejentum-mcp MCP server retrieves task-matched cognitive operations for reasoning, code work, anti-deception checks, and conversation-state perception. Its eight tools send a short task description to the Ejentum API and return a procedure paired with a reasoning or detection topology for the model to absorb before responding. Dynamic tools are available across tiers, while adaptive variants rewrite procedures and topology nodes for the specific task and require Go or Super tier access. Use it when an agent needs structured support for non-trivial analysis, software engineering, integrity-sensitive requests, or observed conversational patterns.

Use cases

  • Improve multi-step analytical answers
  • Review or debug non-trivial code
  • Check responses for deceptive framing
  • Sharpen observations about conversation patterns

Key features

  • Eight MCP retrieval tools
  • Dynamic and adaptive operation modes
  • Structured procedures with reasoning topologies
  • Code-focused engineering operations
  • Anti-deception and perception operations
  • Stdio and hosted HTTP access

Tools8

What Ejentum MCP lets your agent do. Expand a tool for its parameters.

Verified live~3.3k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 3d ago

Recent health check succeeded.

19/19 checks healthy over the last 80d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 9d ago via OSV.dev · ejentum-mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars16Stargazers on the source repository.
npm downloads251Package downloads in the last 30 days.
Last commit4mo agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed8Callable tools this server registers over MCP.
Directory activity6 viewsConfig copies, upvotes, and views on AllMCPs.

How Ejentum MCP works

What ejentum/ejentum-mcp MCP server does

The ejentum/ejentum-mcp MCP server adds eight MCP tools that retrieve structured cognitive operations before an agent produces an answer. Four dynamic tools cover general reasoning, software engineering, anti-deception, and perception of conversation patterns. Four adaptive tools cover the same areas but use an adapter model to rewrite the retrieved operation with identifiers and details from the supplied task.

The returned content is not an answer to the user’s request. It is an instruction payload for the calling model to read and apply internally. A record includes a task-specific procedure, a graph-like topology, cognitive-style guidance, and a verification or perception check. It also identifies a failure pattern to avoid and the expected correct response shape.

How it works

Each tool accepts one query string containing a one- or two-sentence framing of the task. The stdio package and hosted HTTP endpoint forward the request to the same Ejentum upstream API with the selected mode. The API returns an array containing one object whose key matches the requested mode.

Dynamic retrieval selects the top abstract operation by semantic similarity. The four dynamic libraries contain 311 reasoning operations, 128 software-engineering operations, 139 anti-deception operations, and 101 perception operations. Adaptive retrieval considers the top five matches, selects one, and has an adapter LLM rewrite the procedure and topology for the task. Adaptive calls add approximately two to three seconds of latency.

The topology is a structured outline rather than a program executed by the host. It can contain sequential steps, decision gates, parallel branches, bounded loops, negative anchors, meta-cognitive exits, and escape paths. The calling model is expected to absorb both the natural-language procedure and this topology before responding.

Setup and configuration

The ejentum/ejentum-mcp MCP server requires an Ejentum API key and Node.js 18 or newer for the local stdio setup. The documented direct command is:

Terminal
npx -y ejentum-mcp

Configure the command as an MCP subprocess and provide the key through EJENTUM_API_KEY. The README gives configurations for Claude Desktop, Cursor, Windsurf, Claude Code, Cline, Continue, Codex CLI, and n8n’s MCP Client. It also documents a hosted Streamable HTTP endpoint at https://api.ejentum.com/mcp, which uses a bearer token instead of a local subprocess.

Dynamic tools are available on all tiers, including the 30-day free trial. Adaptive tools require the Go or Super tier. Requests use bearer authentication against the upstream API; unauthorized, tier-restricted, and rate-limited requests return corresponding HTTP errors.

Tools and capabilities

  • reasoning: supports analytical, diagnostic, planning, comparison, architecture, and other multi-step questions.
  • code: supports code generation, review, refactoring, debugging, algorithm selection, and dependency-change evaluation.
  • anti-deception: addresses pressure to agree, unsupported certification, authority-based framing, and similar integrity risks.
  • memory: sharpens an observation already formed about conversation state, user behavior, emotional shifts, or cross-turn patterns.
  • adaptive-reasoning, adaptive-code, adaptive-anti-deception, and adaptive-memory: provide task-rewritten versions of the corresponding dynamic operations.

The tools are intentionally not general-purpose lookup, file-reading, code-execution, or structured-memory tools. The supplied task framing should describe what the model is reasoning about, coding, evaluating, or observing.

Limitations and notes

Adaptive modes are gated by subscription tier and can be slower than dynamic retrieval. The memory modes are for sharpening an existing observation, not for extracting facts, summarizing, or storing structured data. The server depends on the Ejentum API and its rate limits, so a valid key and available tier are required for successful requests. The returned injection contains plain text and must be interpreted by the connected model; the topology is not executed by a separate runtime.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Ejentum MCP

Run `npx -y ejentum-mcp` with Node.js 18 or newer, then configure it as an MCP subprocess with `EJENTUM_API_KEY`.

Yes.

It provides dynamic `reasoning`, `code`, `anti-deception`, and `memory` tools, plus adaptive variants of each.

Adaptive modes require the Go or Super tier. Dynamic modes are available on all tiers, including the 30-day free trial.

Related MCP servers

All in Coding Agents All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewEjentum MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/ejentum-ejentum-mcp?style=directory)](https://allmcps.com/mcp/ejentum-ejentum-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/ejentum-ejentum-mcp"><img src="https://allmcps.com/api/badge/ejentum-ejentum-mcp?style=directory" alt="Ejentum MCP on AllMCPs" /></a>