Index HUMI – Agent Human-like Intelligence Index

Version 1.1
Date: July 17, 2026
Designed for: GlobalScoreAgent Ecosystem (ERC-8004)

What is the Index HUMI?

The Index HUMI (Human-like Metrics Index) is a powerful, single-number reputation score ranging from 0 to 100 that measures the overall quality, legitimacy, maturity, and real-world intelligence of any Agent in the ERC-8004 ecosystem.

It is calculated by combining four independent pillars, each worth a maximum of 25 points, for a total of 100 points. Every point is derived from real, verifiable on-chain data (metadata, activity, owner history, external analysis, and protocol usage).

HUMI provides a clear, transparent, and easy-to-understand metric that answers the key question:
“How trustworthy, mature, and human-like is this Agent?”

It is the core reputation layer for every Agent registered in the ERC-8004 network.

HUMI Maturity Levels (Business Interpretation)

The HUMI score (0–100) is translated into 5 clear maturity/confidence levels. These levels are used in the dashboard for:

Filters and advanced search
Badges and colors on Agent cards
Ranking and recommendations
Risk alerts when combined with WAMI

Level	HUMI Range	UI Color	Badge / Label	Description for Users (Dashboard)	Business Trust Level	Approx. Risk
Unstable	0 – 49	🔴 Red	Unstable	High-risk or very low maturity Agent. Extreme caution required.	Very low – Not recommended for production	High
Developing	50 – 64	🟠 Orange	Developing	Basic Agent. Has minimal presence but still immature.	Moderate – Monitor closely	Medium-High
Stable	65 – 79	🟢 Light Green	Stable	Reliable Agent with solid intermediate maturity. Recommended for general use.	Good – Acceptable confidence level	Low
Very Stable	80 – 89	🟢 Green	Very Stable	Mature, consistent, and high-quality Agent. High reliability.	High – Suitable for critical integrations	Very Low
Elite	90 – 100	🟢 Dark Green	Elite	Reference Agent in the ecosystem. Maximum quality and maturity.	Very High / Premium – Maximum trust	Minimal

How Index HUMI Applies to the ERC-8004 Ecosystem

In the ERC-8004 ecosystem, Agents are autonomous entities that interact with users, protocols, and each other. Their credibility is critical for adoption, governance, staking, and marketplace listings.

Index HUMI acts as the primary trust layer for Agents, evaluating not just technical features but the full picture of legitimacy and real activity. It is calculated daily using the official Graph from Ormi Labs together with transactional-wallet activity across the indexed chains. Nonce, balance, and deltas are calculated per wallet and chain, then aggregated by agent.

HUMI applies to all Agents — new or ancient, high-activity or low-activity — ensuring every registered Agent receives a fair, up-to-date reputation score.

It enables:

Automatic ranking and filtering of Agents in dashboards and marketplaces.
Real-time risk assessment before any interaction (executions, attestations, payments, etc.).
Seamless integration with Index WAMI to create a complete reputation system (Agent + Owner wallet).

By adding HUMI, the entire ERC-8004 network gains a standardized way to distinguish high-quality, battle-tested Agents from low-effort or suspicious ones.

Benefits of Using Index HUMI

Instant Trust Decision: One number (0–100) tells users, platforms, and protocols everything they need to know about an Agent’s quality.
Risk Reduction: Automatically flags Agents with poor ownership history, low metadata quality, suspicious activity, or weak external validation.
Better Wallet Reputation Synergy: High-HUMI Agents paired with high-WAMI owner wallets receive maximum trust signals.
Transparent & Auditable: Every point is backed by on-chain data and fully explainable (no black-box AI).
Scalable & Real-Time: Updates automatically as new Agent activity and analysis records are processed.
Business Opportunities: Enables premium features such as Agent scoring APIs, advanced search filters, governance weighting, and marketplace curation.
Ecosystem-Wide Standardization: Creates a common language of trust across all Agents, owners, and dApps in the ERC-8004 network.

How Index HUMI Complements the Index WAMI

Index HUMI and Index WAMI work together as a complete reputation stack for the ERC-8004 ecosystem:

Index HUMI evaluates the Agent itself (metadata quality, activity, identity, protocol usage, owner history, etc.).
Index WAMI evaluates the wallet behind the Agent (the owner or registering wallet).

Key synergy:

A high HUMI score (e.g., 85+) is further strengthened when the owner wallet also has a high WAMI score, creating the strongest possible trust signal.
A low HUMI score (e.g., below 60) combined with a low WAMI score raises strong red flags, even if one of them looks acceptable in isolation.
Together they provide the full picture: “Is the Agent well-built AND is it controlled by a trustworthy wallet?”

This combination creates the most robust reputation system in the ERC-8004 ecosystem — far stronger than evaluating Agents or wallets in isolation.

Comparison with External Reputation Indices and Oracles

Index / Oracle	Provider	Focus	Score Range	Data Used	Key Advantage of HUMI
Nansen / Arkham Agent Scores	Nansen / Arkham	Wallet & entity labeling	0–100	Off-chain + on-chain analytics	HUMI is purpose-built for ERC-8004 Agents
Chainalysis / TRM Labs	Chainalysis/TRM	Risk & compliance scoring	Risk tiers	On-chain + off-chain intel	HUMI is public, transparent, and Agent-specific
Dune / Community Dashboards	Open-source	Custom on-chain metrics	Varies	On-chain queries	HUMI is standardized, real-time, and integrated with GSA
EigenLayer / DeFi Reputation	Various DeFi	Staking & protocol reputation	Varies	Protocol-specific activity	HUMI evaluates the full Agent lifecycle with meaningful multichain presence
General AI Agent Scores	Various startups	Off-chain AI metrics	Varies	API + metadata	HUMI is fully on-chain and tied to ERC-8004 registrations

Why HUMI stands out:

It is purpose-built for the ERC-8004 Agent ecosystem (most external tools are general-purpose or wallet-only).
Fully on-chain and transparent (no proprietary black-box models).
Directly integrated with Index WAMI for a complete Agent + Owner reputation system.
Free from external dependencies — all data is sourced directly from the blockchain (ERC-8004 Graph from Ormi Labs + wallet activity indexed via Alchemy, Moralis, and Zerion) and stored in our ERC-8004 and walcert tables.

Data Freshness and Agent Evaluation Strategy

Index HUMI is engineered for both accuracy and efficiency. It is recalculated daily using the official Graph provided by Ormi Labs together with transactional-wallet activity, ensuring every Agent has fresh, reliable data.

Which Agents are evaluated?

HUMI evaluates all Agents in the ecosystem without exception:

New Agents and ancient Agents
High-activity Agents and low-activity Agents
Agents whose transactional activity is consolidated across valid wallets and indexed chains

This universal approach guarantees that every registered Agent receives a fair and current reputation score.

What data is refreshed daily?

Nonce, balance, and deltas calculated per transactional wallet and chain, then aggregated by agent.
On-chain activity and registration data across the indexed chains, with emphasis on presence quality rather than simply counting how many chains an agent appears on.
Off-chain pointers found directly in ERC-8004 records, including:
- Metadata URIs and DID documents
- Feedbacks from more than 21 external entities (processed automatically)

The system intelligently decides which modules need updating based on the latest Graph data and time-based freshness rules. This daily cycle, combined with smart “does need” flags, keeps the index accurate without unnecessary recomputation.

Benefits of this approach

Maximum freshness: Scores are never more than 24 hours old.
Complete coverage: Every Agent — regardless of age or activity level — is included.
Rich data depth: Combines on-chain Graph data with imported off-chain metadata, external feedbacks, and multichain aggregation of valid transactional wallets.
Scalability: Daily processing across indexed chains remains efficient thanks to intelligent filtering.
Trust at ecosystem scale: Users and platforms always see the most current and comprehensive view of Agent quality.

This strategy ensures Index HUMI remains both accurate and performant at ecosystem scale.

Pillars of the Index HUMI

Each pillar is designed around a nominal maximum of 25 points. Below are the specific aspects analyzed and how they contribute to the score.

1. Pillar History (25 points)

Evaluates the Agent’s ownership stability and historical reputation.

Strength and activity of the owner wallet: 10 points
Stability of ownership over time (few or no changes): 5 points
Longevity and antiquity of the owner wallet: 5 points
Quality and activity of the owner’s overall Agent portfolio: 5 points

2. Pillar Information (25 points)

Measures the richness, professionalism, and completeness of the Agent’s public identity and technical metadata.

Quality of name, description, and image: 7.5 points
Diversity and depth of information sources (Chain + URI + external): 7.5 points
Availability of contact methods and programmatic endpoints: 5 points
Advanced technical maturity (supported trust, verification methods, skills, etc.): 5 points

3. Pillar Measure (25 points)

Evaluates whether an agent is credibly represented, externally validated, and operationally present.

Metadata richness and operational existence (aggregated nonce, attestations, executions, or feedback)
External validation, identity analysis, and protocol quality
Multichain Presence Quality — meaningful depth in valuable and high-value chains, not superficial presence on more chains
Warning penalties from agent alerts

Transactional wallet volume still contributes, but with reduced direct weight; quality of multichain presence is now an explicit scoring dimension.

4. Pillar Usage (25 points)

Evaluates recent operational adoption across wallets and chains.

Recent aggregated wallet activity from all valid transactional wallets
On-chain channel quality (attestations, executions, feedbacks, protocol activity)
Comments and paid protocol usage
Multichain Valuable Usage — meaningful activity across valuable chains
Multichain High Value Consistency — presence in high-value chains without extreme concentration on a single chain
Revocation penalties for recent revoked activity

Appearing on more chains by itself is not rewarded. Usage rewards valuable multichain activity and penalizes shallow spread or extreme primary-chain concentration.

Ready to implement?
For the technical implementation of the API and access to the Dashboard, we recommend visiting our official website:
https://www.globalscoreagent.com/

Together with Index WAMI, HUMI forms the new standard for trust in the ERC-8004 ecosystem.

Documentation