SkillAgentSearch skills...

trustlayer-sybil-scanner

Feedback forensics for ERC-8004 agents. Detects Sybil rings, fake reviews, rating manipulation, and reputation laundering across 20 chains. No API key needed.

Install / Use

npx skills add BankrBot/skills --skill trustlayer-sybil-scanner

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

92/100

Category

Security

Supported Platforms

Universal

Our assessment of trustlayer-sybil-scanner

trustlayer-sybil-scanner scores 92/100 on our quality scale, 401st of 1,108 Security skills we index (top 37%).

Its SKILL.md is 7.1 KB long, well organised into 11 sections with 11 code examples: a thorough specification that gives an agent plenty to work with.

With 1,201 GitHub stars, it is one of the more widely adopted skills in the catalogue.

Substance
29/30
Structure
20/20
Description
15/15
Adoption
13/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so trustlayer-sybil-scanner is actively maintained.
  • Our last check on 2026-10-28 found the source still online.
  • No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
  • Its trust signals score 88/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-10-07. Automated pattern scan on 2026-10-07. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

trustlayer-sybil-scanner compared with similar skills

All 4 of these similar skills score higher than trustlayer-sybil-scanner; compare them before choosing.

SkillScoreStarsUpdatedFormat
trustlayer-sybil-scanner (this skill)by BankrBot921.2k54d agoSKILL.md
Agent-Reachby Panniantong100112.3k8d agoCLAUDE.md
headroomby headroomlabs-ai10079.8ktodayCLAUDE.md
Scraplingby D4Vinci100103.6k2d agoMCP Server
crawl4aiby unclecode10087.4k3d agoMCP Server

Frequently asked questions

How do I install trustlayer-sybil-scanner?
Run npx skills add BankrBot/skills --skill trustlayer-sybil-scanner. The install tabs above show the steps for each supported agent.
Which AI agents does trustlayer-sybil-scanner work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is trustlayer-sybil-scanner safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It declares no license and scores 88/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is trustlayer-sybil-scanner still maintained?
The repository was last updated about 2 months ago, so trustlayer-sybil-scanner is actively maintained.

name: trustlayer-sybil-scanner description: Feedback forensics for ERC-8004 agents. Detects Sybil rings, fake reviews, rating manipulation, and reputation laundering across 20 chains. No API key needed. version: 4.1.0 tags:

  • reputation
  • trust
  • sybil
  • erc-8004
  • x402
  • security
  • agents metadata: clawdbot: emoji: "🔍" homepage: "https://thetrustlayer.xyz" requires: bins: - curl

TrustLayer Sybil Scanner — ERC-8004 Feedback Forensics

Detects fake reviews, Sybil rings, rating manipulation, and reputation laundering in ERC-8004 agent ratings across 20 chains including Base, Ethereum, BSC, Celo, Gnosis, Optimism, Arbitrum, Avalanche, Polygon, Monad, and Solana.

Most agent reputation systems show you the rating. This one tells you if the rating is real.

API Base: https://api.thetrustlayer.xyz No API key required (beta). x402 micropayments on paid endpoints ($0.001 USDC per query).

What this catches that others don't

  • Sybil rings: Clusters of wallets that only review each other
  • Reviewer quality: A 5-star review from a trusted agent vs a throwaway wallet are not the same. Scores are weighted by reviewer reputation.
  • Cross-chain laundering: Agent has great reviews on BSC but terrible ones on Base? Identities are resolved across chains and score divergence is flagged.
  • Temporal anomalies: Sudden bursts of positive reviews after a period of bad ones. Review bombing. Rating manipulation patterns over time.
  • Spam feedback: Known spam patterns detected via tag analysis and filtered before scoring.

When to use this skill

  • Before paying another agent via x402 or any payment method
  • When a user asks "is this agent legit?" or "are these reviews real?"
  • Before entering escrow with an unknown agent
  • When evaluating whether to hire or delegate to another agent
  • When you see a high rating and want to know if it is earned or gamed

Sybil Scan

curl -s "https://api.thetrustlayer.xyz/trust/<chain>:<agentId>"

Replace <chain> with: base, ethereum, bsc, celo, gnosis, optimism, arbitrum, avalanche, polygon, monad, mantle, scroll, linea, metis, taiko, solana-mainnet (and more) Replace <agentId> with the numeric agent ID.

Example:

curl -s "https://api.thetrustlayer.xyz/trust/base:1378"

What comes back

The response is JSON. The fields that matter most for fraud detection:

Sybil & fraud signals:

  • sybil_risk: "low" / "medium" / "high". High means the reviewer network looks artificial.
  • anomaly_flags: Array of detected issues. Empty = clean. Watch for: review_bombing, duplicate_feedback_content, single_agent_reviewers, spam_feedback, reputation_laundering.
  • reviewer_weighted_score: The trust score after adjusting for reviewer credibility. If this is much lower than trust_score, the agent's good reviews are coming from low-quality reviewers.

Cross-chain signals:

  • cross_chain_scores: If present, this agent exists on multiple chains. Check laundering_risk and score_divergence. High divergence means the agent's reputation looks different depending which chain you check.

Overall assessment:

  • trust_score (0-100): Composite score. 80+ low risk, 64-79 medium, below 64 high risk.
  • risk_level: "low" / "medium" / "high". Quick decision signal.
  • recommended_max_exposure_usd: How much USD to risk with this agent.
  • confidence: "low" / "medium" / "high" based on data volume.

Decision logic

IF sybil_risk == "high"           → BLOCK. Tell the user why.
IF "review_bombing" in flags      → BLOCK. Likely manipulated.
IF "reputation_laundering" in flags → BLOCK. Cross-chain fraud pattern.
IF risk_level == "high"           → BLOCK. Overall high risk.
IF "spam_feedback" in flags       → WARN. Rating may be inflated.
IF cross_chain laundering_risk == "high" → WARN. Check each chain.
IF reviewer_weighted_score < trust_score - 15 → WARN. Good ratings from bad reviewers.
IF recommended_max_exposure < tx amount → REDUCE exposure or warn.
ELSE → proceed.

Pre-payment check example

About to pay agent base:5000 for a $50 service:

RESULT=$(curl -s "https://api.thetrustlayer.xyz/trust/base:5000")

# Extract key fraud signals
SYBIL=$(echo "$RESULT" | grep -o '"sybil_risk":"[^"]*"' | cut -d'"' -f4)
RISK=$(echo "$RESULT" | grep -o '"risk_level":"[^"]*"' | cut -d'"' -f4)
SCORE=$(echo "$RESULT" | grep -o '"trust_score":[0-9]*' | cut -d':' -f2)
FLAGS=$(echo "$RESULT" | grep -o '"anomaly_flags":\[[^]]*\]')

Report to user: "Scanned base:5000. Trust score: $SCORE. Sybil risk: $SYBIL. Anomaly flags: $FLAGS"

If sybil_risk is high: "This agent's reviews show signs of Sybil manipulation. Recommend not transacting."

Other endpoints

Agent lookup (paid $0.001 USDC — returns full agent profile, metadata, and on-chain registration details):

curl -s "https://api.thetrustlayer.xyz/agent/<chain>:<agentId>"

Leaderboard (most trusted agents, Sybil-filtered — rate-limited: 5 free per IP per hour, then 402):

curl -s "https://api.thetrustlayer.xyz/leaderboard?chain=base&limit=10"

Network stats (live counts of total agents, Sybil flags, chains covered, and more):

curl -s "https://api.thetrustlayer.xyz/stats"

Reviewer lookup (paid $0.001 USDC — returns reviewer quality score, total reviews, unique agents reviewed, quality tier, and recent review history):

curl -s "https://api.thetrustlayer.xyz/reviewer/<wallet_address>"

Most reviewers score low — a trusted reviewer is rare. Use this to verify if a reviewer is credible before trusting their feedback.

Owner portfolio (paid $0.001 USDC — returns all agents owned by one wallet across chains, with cross-chain group info, average trust score, and risk assessment):

curl -s "https://api.thetrustlayer.xyz/owner/<wallet_address>"

Use for due diligence on an agent operator.

Score history (paid $0.001 USDC — returns full daily score time-series, 7d/30d trajectory, and volatility):

curl -s "https://api.thetrustlayer.xyz/history/<chain>:<agentId>"

Daily snapshots retained for 90 days. Use to check if an agent's reputation is stable or volatile.

Call /stats for current network coverage — agent counts, Sybil flags, cross-chain groups, and chain breakdown are all returned live.

Visual reports

For a full visual breakdown with score history, anomaly timeline, and cross-chain map:

https://thetrustlayer.xyz/agent/<chain>:<agentId>

How scoring works

Scores combine three dimensions, each weighted by data quality:

  1. Profile completeness: Does the agent have metadata, description, active endpoints?
  2. Feedback volume: How much feedback exists? Weighted by reviewer quality, not raw count.
  3. Feedback legitimacy: Are reviewers themselves reputable? Are there Sybil patterns? Spam? Temporal anomalies?

Six Sybil detection methods run on every sync:

  • Reviewer overlap clustering
  • One-to-one review pattern detection
  • Wallet age and activity analysis
  • Cross-chain identity correlation
  • Feedback timing anomaly detection
  • Tag-based spam filtering

Scores update daily. Historical score snapshots retained for 90 days.

Related Skills

View on GitHub
GitHub Stars1.2k
CategorySecurity
Updated1mo ago
Forks629

Languages

JavaScript

Trust signals

88/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 medium