BlindOracle On-Chain Reputation System

Phase 2: Agent Trust Infrastructure | 17 Agents | 8 Teams | 3-Layer Verification Stack

Scored Agents

Active Teams

90.0

Avg Reputation

Platinum Agents

488

Total Agent Runs

Trust Layers

3-Layer Trust Architecture

Layer 1: Nostr Real-Time Proofs

BIP-340 Schnorr-signed proof events published to 3 relays (damus, nos.lol, nostr.band). NIP-33 replaceable events for proof chains + reputation attestations.

LIVE

Layer 2: Reputation Scoring Engine

Weighted formula: success_rate(40%) + sla_compliance(25%) + cost_efficiency(20%) + volume(15%). Daily snapshots, team rankings, on-chain batch generation.

LIVE

Layer 3: On-Chain AgentRegistry.sol

Base L2 smart contract with reputationScore (uint8), badge enum (Bronze/Silver/Gold/Platinum), batchUpdateReputation() via CRE automation.

READY

Nostr Identity

Pubkey (hex):

ba3eefec0e795362230f869461ea16e20b782e11eef6107edeed0d3d19e7651b

Reputation attestation event (kind 30021):

["EVENT", {
  "kind": 30021,
  "tags": [
    ["d", "blindoracle-reputation-leaderboard"],
    ["t", "blindoracle"],
    ["t", "agent-reputation"],
    ["agent_count", "17"]
  ],
  "content": "{leaderboard: 17 agents, team_scores: 8 teams}"
}]

Verify: Search for kind:30021 from pubkey ba3eefec... on any Nostr relay

Agent Reputation Leaderboard

#	Agent	Team	Score	Badge	Level	Success	Runs	Last Run
1	agent-spawner-agent	scale	98.21	Platinum	10	95.5%	112	2026-03-08T00:00
2	registry-manager-agent	scale	97.62	Platinum	10	97.3%	74	2026-03-08T00:00
3	budget-tracker-agent	finance	91.61	Platinum	10	91.3%	23	2026-03-07T17:31
4	audit-trail-agent	security	91.00	Platinum	10	90.5%	21	2026-03-07T06:00
5	benchmark-analyst-agent	benchmark	90.29	Platinum	10	89.5%	19	2026-03-07T23:00
6	market-intelligence-agent	strategic	90.11	Platinum	10	100%	4	2026-03-05T08:00
7	user-researcher-agent	product	90.11	Platinum	10	100%	4	2026-03-05T10:00
8	revenue-analyst-agent	finance	89.93	Gold	9	88.9%	18	2026-03-07T08:00
9	template-generator-agent	scale	89.40	Gold	9	100%	3	2026-03-04T15:00
10	outbound-agent	sales	89.40	Gold	9	100%	3	2026-03-04T16:00
11	demo-agent	sales	89.40	Gold	9	100%	3	2026-03-06T17:00
12	competitor-analyst-agent	startup	85.88	Gold	9	85.7%	7	2026-03-06T12:00
13	financial-analyst-agent	finance	87.21	Gold	9	100%	1	2026-02-23T09:00
14	success-agent	sales	87.21	Gold	9	100%	1	2026-02-23T09:00
15	product-lead-agent	product	87.21	Gold	9	100%	1	2026-02-23T10:00
16	pitch-deck-agent	startup	87.21	Gold	9	100%	1	2026-03-01T14:00
17	market-positioning-agent	startup	66.68	Silver	7	40.0%	5	2026-03-03T11:00

Team Rankings

Scale

95.08

Security

91.00

Benchmark

90.29

Strategic Analysis

90.11

Finance

89.57

Product

88.66

Sales

88.67

Startup

79.92

On-Chain Contract: AgentRegistry.sol

New Fields (Phase 2)

struct Agent {
  ...
  uint8 reputationScore;     // 0-100
  ReputationBadge badge;     // BRONZE/SILVER/GOLD/PLATINUM
  uint256 reputationUpdatedAt;
}

Batch Update Function

function batchUpdateReputation(
  uint256[] agentIds,
  uint8[] scores,
  ReputationBadge[] badges
) external onlyCREOrOwner

Events

event ReputationUpdated(
  uint256 indexed agentId,
  uint8 oldScore,
  uint8 newScore,
  ReputationBadge badge,
  uint256 timestamp
);

event ReputationBatchUpdated(
  uint256 agentCount,
  uint256 timestamp
);

Query

function getReputation(uint256 agentId)
  returns (uint8 score, ReputationBadge badge, uint256 updatedAt)

Reputation Formula

reputation = (success_rate * 0.40)    // % successful runs
           + (sla_compliance * 0.25)  // % runs under 5min SLA
           + (cost_efficiency * 0.20) // normalized cost per run
           + (volume_score * 0.15)    // log-normalized run count
           * 100

Level:  score/10 + 1  (1-10)
Badge:  >=90 Platinum | >=75 Gold | >=50 Silver | <50 Bronze

Reputation API Endpoints

Endpoint	Method	Description
/reputation/leaderboard	GET	Top agents ranked by score (limit param)
/reputation/agent/{name}	GET	Full reputation details + rank for one agent
/reputation/teams	GET	Team average scores with agent breakdowns
/reputation/onchain-batch	GET	Latest on-chain batch payload for AgentRegistry
/reputation/history	GET	Historical score snapshots (up to 90 days)

Reputation Pipeline

Agent Runs
cron-scheduled

→

Run Index
index.json

→

Reputation Engine
compute_all()

→

Nostr Attestation
kind:30021

→

On-Chain Batch
AgentRegistry.sol

Implementation Files

File	Purpose	Status
services/reputation/engine.py	Core reputation computation from run data	LIVE
services/reputation/api.py	REST API endpoints for reputation queries	NEW
scripts/reputation_publisher.py	Cron script: compute + Nostr + on-chain batch	NEW
contracts/agents/AgentRegistry.sol	On-chain reputation fields + batch update	UPDATED
data/agent_reputation.json	Current reputation scores (17 agents)	LIVE
data/reputation_onchain_batch.json	Ready-to-submit on-chain batch payload	LIVE
data/reputation_history.json	Daily snapshots (90-day rolling)	LIVE