Distributed Truth, Human-Centric Intelligence
At present, the DXT CLI may report 'Successfully signed' but fail to verify signed extensions. This is a documented CLI/tool issue and does not affect installation for most users. Signature verification will be fully supported as soon as upstream tool updates are released.
MCP (Multi-Agent Collaboration Protocol) is an open, extensible framework for collaborative, decentralized truth validation and resilient evidence networks. Agents—AI or human-augmented—work independently but converge on verifiable, tamper-evident knowledge through standardized protocols.
- AGENTS.md: Universal BIAS standards for context-aware, bias-resistant agent analysis and reporting.
- MCP.md: The protocol for agent publication, verification, and network-wide audit.
- Humanify Template: Ensures all agent/AI answers center human values, attribution, and ethical safeguards.
Join the MCP/BIAS Trust Network!
- Contribute your verified truths
- Curate, fork, and vet evidence alongside fellow explorers
- Empower both people and AI with transparent, distributed knowledge
Your participation makes collective intelligence a reality—start by sharing your own truth, or explore trusted records from the community!
- For Humanify (AI/LLM prompt integration), start below.
- For agent developers or node operators, see AGENTS.md and MCP.md.
“Paste it anywhere. Launch a node.
Truth, distributed, is unkillable.”
Humanify Template is a drop-in prompt framework designed to help any Large Language Model (LLM)—like GPT, Claude, or custom AI bots—return answers that are fair, accurate, responsible, and centered on human values.
It does so by guiding the model to:
- Attribute responsibility at the system level
- Highlight missing or underrepresented perspectives
- Encourage ethical/social framing and feedback loops
Whether you're building custom AIs or just want better answers from existing ones, this template works for you!
- Get the Humanify Template:
View and copy the template here - Paste it into the system prompt, custom instructions, or configuration box of:
- Claude Apps
- GPT custom bots
- Any AI chat or workflow with a prompt template field!
- Send your question as usual and enjoy answers that are more fair, comprehensive, and human-centric.
Visit humanify-template.md to copy the latest version.
On GitHub, the code block includes a clipboard icon for one-click copying.
If using locally, select all the text inside the code block and copy it!
- Fork or clone this repo to add your own templates, improvements, or prompt tools.
- Add om-design/MCP as a resource in your own AI/LLM or plugin documentation.
- Pull requests and prompt innovation are always welcome!
- Open Issues for suggestions/bugs.
- PRs for new templates, translations, or workflow integrations are welcome!
- Help keep this project improving for all.
Share this repo with anyone seeking fairer, more ethical AI answers. Every star, fork, and share helps bring human values to AI’s core reasoning!
© 2025 omdesign. All rights reserved.
LICENSE DETAILS: https://github.com/om-design/MCP/blob/main/LICENSE.md
-
Personal or research use:
You may copy and use the Humanify Template in your own LLMs, Claude Apps, GPTs, or projects with clear attribution ("Humanify Template by om-design/MCP" + link to this repo). -
Attribution Required:
Any use, distribution, or reference must include clear credit and a link back to om-design/MCP.
📝 Suggested Attribution Example
"Powered by the Humanify Template (© omdesign, GitHub link)"
-
No Unapproved Forking or Redistribution:
Forking, extensive modification, creating variants/templates for redistribution, or publishing this template in other repositories or commercial projects requires written permission from om-design. -
Contributions Welcome:
You can propose improvements, open issues, or submit pull requests for positive changes. All accepted contributions will credit their authors. -
Misuse or Abuse: We will not tolerate any action by any party that, in our judgment, diminishes the agency of any human.
📬 For permission requests, misuse notifications, suggestions, or other inquiries,
please open an Issue on GitHub.