Skip to main content

Elemm: The Landmark Manifest Protocol

PyPI version Downloads Build Status License Discord Python versions Website

Elemm Logo

The Infrastructure for the Agentic Web.

Elemm is the Landmark Manifest Protocol, a next-generation communication framework designed to transform how autonomous LLM agents interact with the digital world. Instead of static tool definitions, Elemm provides a dynamic, manifest-driven architecture that enables agents to discover, navigate, and execute complex workflows across distributed APIs with unprecedented efficiency.


Quick Start

1. Install

python3 -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate
pip install elemm

2. Generate your MCP Client Configuration

The easiest way to connect your MCP Client (e.g., Cursor, Claude Desktop) is to use the built-in configuration generator. It automatically detects your OS, paths, and WSL environment:

elemm-gateway cfg-gen

Follow the interactive prompts to generate your JSON config snippet, then copy and paste it into your MCP client's configuration file.

3. Start the Dashboard (Optional but Recommended)

Launch the interactive dashboard to monitor tokens, manage credentials (Vault), configure security policies, and generate configs visually:

elemm-dashboard

Open http://localhost:8090 in your browser.

4. Start Discovering

Once connected, tell your agent:

"Use Elemm to connect to https://petstore.swagger.io/v2/swagger.json and list all available pets."

The Gateway provides 9 core tools to the agent. All domain-specific actions are discovered on-the-fly via the Elemm protocol.

5. Build Your Own Landmark Server (Optional)

You can turn any Python function into a high-performance landmark using decorators. Depending on your needs, you can expose these landmarks in two ways:

Option A: FastAPI (Web-based via HTTP/SSE)

Exposes the landmarks as a standard web service / API:

import uvicorn
from fastapi import FastAPI
from elemm import AIProtocolManager, MetadataRegistry
from elemm.gateways.fastapi import FastAPIGateway

app = FastAPI()
registry = MetadataRegistry("landmarks.yaml")
manager = AIProtocolManager(registry=registry)

@manager.landmark("security:quarantine_node")
async def quarantine_node(node_id: str, urgent: bool = False):
    """Quarantines a compromised server node."""
    return {"status": "success", "node": node_id}

# Expose the Elemm manifest and endpoint dynamically on your FastAPI app
gateway = FastAPIGateway(manager)
gateway.bind_to_app(app)

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=8000)

Option B: Standalone MCP (Local via STDIO)

Exposes the landmarks directly as a local STDIO MCP server:

from elemm import ElemmGateway

# Initialize the high-level gateway wrapper
gateway = ElemmGateway(name="MySecurityServer")

@gateway.action("security:quarantine_node")
async def quarantine_node(node_id: str, urgent: bool = False):
    """Quarantines a compromised server node."""
    return {"status": "success", "node": node_id}

# Run directly as a local STDIO MCP Server!
if __name__ == "__main__":
    gateway.run_mcp()

Advanced Usage

  • Pydantic Discovery: Elemm automatically generates schemas from Pydantic models.
  • Response Hygiene: Built-in _select, _filter, _limit, and _offset parameters prevent context overflow.
  • Session Isolation: Use session_id to run parallel tasks without cross-contamination.
  • Self-Healing: The SmartRepair engine provides agents with actionable remedies when errors occur.
  • Search: Use search_landmarks(query) with Python REGEX to locate tools instantly without full hierarchy traversal.
  • Dashboard: Start the dashboard via elemm-dashboard for a real-time observability UI on port 8090.

The Vision: Agentic Web

In the Agentic Web, every API is a "Landmark". Agents no longer need massive, hardcoded system prompts to understand a service. They discover capabilities on-the-fly via a standardized manifest, just like a human navigates a website.

  • Unified Discovery: Every Elemm-compliant server exposes its structure at /.well-known/elemm-manifest.md.
  • Zero System Prompt: By providing rich semantic landmarks and manifest-driven discovery, you can eliminate thousands of tokens from your system prompts. The protocol is the documentation.
  • One MCP Server, Infinite APIs: The built-in Elemm Gateway connects to any OpenAPI, GraphQL, or native Elemm service. A single pip install gives you a universal MCP server that discovers and loads landmarks on-the-fly, allowing you to scale your agent's capabilities without ever restarting your infrastructure.

The Philosophy: Decoupling Intelligence

Elemm is more than just a protocol; it's a shift toward Decentralized Intelligence. In the traditional SaaS model, providers often bundle their APIs with expensive, centralized LLM interfaces. Elemm decouples the "Body" (the API) from the "Brain" (the Agent).

Bring Your Own Agent (BYOA)

With Elemm, API providers only define the Landmarks and Manifests. The user brings their own autonomous agent to the platform. This shifts the computational burden and cost of "reasoning" to the edge—the user's own system.

Sustainability & Efficiency

By eliminating the need for massive, repetitive system prompts and context-heavy tool injections, Elemm significantly reduces the global token footprint of AI interactions.

  • Lower Latency: No more waiting for centralized "gatekeeper" models to process 20k tokens of documentation.
  • Reduced CO2 & Energy: Fewer tokens mean less GPU compute time, directly translating into a lower carbon footprint for every autonomous task.
  • Cost Sovereignty: Providers save on LLM hosting and token costs, while users get the freedom to choose the model that best fits their task and budget.

Core Advantages

Standard protocols like MCP often struggle with large-scale toolsets. Elemm provides a structural solution:

  • Efficient Discovery: Agents only see a high-level manifest, loading detailed tool schemas only when needed (on-demand inspection).
  • Direct Search: search_landmarks(query) enables regex-based tool discovery without traversing the full hierarchy — ideal for large API surfaces.
  • Atomic Sequencing: Execute multiple tool calls in a single LLM turn with native variable piping ($step0.id).
  • Multi-Protocol Gateway: Connect to any OpenAPI, GraphQL, or native Elemm service through a single MCP server.
  • Security Policy Engine: Built-in Guardian mode with Zero-Trust whitelist, pattern blacklists, landmark restrictions, HTTP method filtering, and Data Loss Prevention.
  • SmartRepair Engine: Built-in error handling that provides agents with actionable remedies instead of cryptic stack traces.
  • Token Economy: Reduces input tokens by up to 90% in complex forensic and administrative scenarios.
  • Observability Dashboard: Optional web UI for real-time monitoring, API exploration, and configuration management.

Documentation


License & Commercial Use

Copyright (C) 2026 Marc Stöcker.

Licensed under the Business Source License 1.1 (BSL 1.1).

What does this mean for you?

  • Startups, Indie Developers & Solo Creators: If your organization has fewer than 50 employees (full-time equivalents) AND less than $5,000,000 USD in annual gross revenue, you can use Elemm (both Core and Gateway) completely free of charge for any purpose, including commercial production.
  • Large Enterprises: If you exceed either of these thresholds, you must obtain a commercial license. Please reach out to v3rm1ll1on@proton.me to discuss commercial licensing.
  • Long-Term Open Source Guarantee: 4 years after the release of each version, the license for that version automatically converts to the Mozilla Public License 2.0 (MPL 2.0), ensuring Elemm remains free and open-source in the long run.

For the full license terms, please see the LICENSE file.

Metadata

Release files for elemm 1.3.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for elemm 1.3.1
File Size Uploaded
elemm-1.3.1.tar.gz 569.0 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for elemm 1.3.1
File Interpreter ABI Platform
elemm-1.3.1-py3-none-any.whl Python 3 none any Details

Total release size: 1.1 MB

Release files / elemm-1.3.1.tar.gz

Download URL elemm-1.3.1.tar.gz
Size 569.0 kB
Tags Source
SHA-256 checksum
How to use checksums
7fcb15cd8880397b207f8216f6f738935617c23aba2ad9239c9dd09cbb06c6c2
BLAKE2b-256 checksum
How to use checksums
9bf839cd1c8c1a2cdd9e7c47eb4843fe7d39e810e9232120ba955b39911d5686
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/6.1.0 CPython/3.13.12

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on May 28, 2026.

Transparency log

Release files / elemm-1.3.1-py3-none-any.whl

Download URL elemm-1.3.1-py3-none-any.whl
Size 544.6 kB
Tags Python 3
SHA-256 checksum
How to use checksums
b55584369dc40eaf972ef9a7ed389d0216dfa0ef1918007072c61c59eed4da01
BLAKE2b-256 checksum
How to use checksums
c7f8d9a363f2e18cc4e3c80e6978304e81037818e29c5e3cbe19ec796f46b519
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/6.1.0 CPython/3.13.12

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on May 28, 2026.

Transparency log

Release history Release notifications | RSS feed

This release

1.3.1 This release

2 release files

1.3.0

2 release files

1.2.0

2 release files

1.1.4

2 release files

1.1.3

2 release files

1.1.2

2 release files

1.1.1

2 release files

1.1.0

2 release files

1.0.3

2 release files

1.0.2

2 release files

1.0.1

2 release files

1.0.0

2 release files

0.9.3

2 release files

0.9.2

2 release files

0.9.1

2 release files

0.9.0

2 release files

0.8.1

2 release files

0.8.0

2 release files

0.7.0

2 release files

0.6.0

2 release files

0.5.4

2 release files

0.5.3

2 release files

0.5.2

2 release files

0.5.1

2 release files

0.5.0

2 release files

0.4.0

2 release files

0.3.11

2 release files

0.3.1

2 release files

0.3.0

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page