Metadata Richness Score: Understanding Agent Quality
Overview
The Metadata Richness Score is a 0–100 quality metric that evaluates how complete, professional, and sophisticated an agent's public profile is.
It measures the depth and professionalism of the information an agent presents to the world — from basic identity details to advanced technical capabilities.
A higher score signals that the agent is:
- Easier to discover and understand
- More trustworthy for users and developers
- Better prepared for real-world adoption, integrations, and governance
This score is one of the key components used in the Index Humi (the overall agent reputation index) and directly influences how prominently an agent appears in searches, rankings, and recommendations.
The score is calculated across three progressive layers:
- Basic Layer (up to 40 points) — Core identity and discoverability
- Intermediate Layer (up to 30 points) — Professional maturity and usability
- Advanced Layer (up to 30 points) — Technical sophistication and innovation
The final score is the sum of all three layers (capped at 100).
Layer 1: Basic Layer (0–40 points)
This layer evaluates whether the agent has a solid, professional foundation that allows users to immediately understand what it is and what it does.
| Item | Business Explanation | Maximum Points |
|---|---|---|
| Name | A clear, meaningful name that properly identifies the agent (no generic placeholders) | 12 |
| Description | A detailed, informative description that explains the agent's purpose and value | 12 + up to 4 bonus |
| Image | A professional visual identity (logo or avatar) that represents the agent | 8 |
| Tags | Relevant categorization tags that improve discoverability and search relevance | 8 + up to 2 bonus |
Why this layer matters: Without strong basic metadata, even the most powerful agent is difficult to find and understand. This layer represents the "first impression" that users and applications see.
Layer 2: Intermediate Layer (up to 30 points)
This layer evaluates the agent's professional maturity — how ready it is for real-world use, integration, and trust.
| Item | Business Explanation | Maximum Points |
|---|---|---|
| Verification Methods | Proof that the agent is legitimately controlled and verifiable | 6 |
| Supported Trust Signals | Certifications, attestations, or trust frameworks the agent supports | 5 |
| Services | Clearly defined services/endpoints the agent offers to users and other agents | 6 + up to 4 bonus |
| Payment Capabilities | Ability to accept payments or participate in economic transactions | 5 |
| Governance | Transparency about ownership, control, and decision-making structure | 4 |
Why this layer matters: It shows that the agent is not just "declared" but is professionally set up for actual usage, monetization, and accountability.
Layer 3: Advanced Layer (up to 30 points)
This layer evaluates the agent's technical depth and innovation — how advanced and future-proof its capabilities are.
| Item | Business Explanation | Maximum Points |
|---|---|---|
| OASF Compatibility | Adoption of open, standardized agent frameworks for better interoperability | 8 |
| Technical Tools | Advanced tooling and execution capabilities | 7 + up to 4 bonus |
| Technical Prompts | Sophisticated prompt engineering and interaction capabilities | 7 |
| Technical Capabilities | Rich set of programmable features and behaviors | 8 |
Why this layer matters: It distinguishes truly sophisticated, production-ready agents from basic ones. High scores here indicate the agent can handle complex workflows, integrate deeply with other systems, and support advanced use cases.
How the Total Score is Interpreted
| Score Range | Interpretation | Business Implication |
|---|---|---|
| 85–100 | Excellent / Production-Ready | Top-tier agent with full professional presence |
| 70–84 | Strong / Well-Developed | Reliable and ready for most real-world uses |
| 50–69 | Moderate / Basic | Functional but needs improvement in key areas |
| Below 50 | Limited / Incomplete | Needs significant work to be competitive |
Why the Metadata Richness Score Matters
- For users: Helps quickly identify high-quality, trustworthy agents.
- For developers: Makes it easier to find agents that are integration-ready.
- For the ecosystem: Encourages creators to provide complete, professional metadata, raising the overall quality of the agent network.
- In Index Humi: Contributes directly to the agent's overall reputation and visibility.
The Metadata Richness Score is recalculated whenever new profile information is added or updated, ensuring the score always reflects the agent's current state of professionalism.
Document generated from the official metadata richness calculation logic (May 2026).
For questions about how your agent can improve its score, refer to the detailed layer explanations above or contact the platform support team.
Let me know if you want any section expanded, rephrased, or if you'd like to add examples or recommendations for improving each layer!