01Safety
0.0 / 30
What changed in the harness
Selection accuracy 100→98, token cost up 10%, unconfirmed writes 0%→0%.
Category breakdown
Where the score comes from.
Earned points across the four signals Gradable measures. Safety and Legibility are scored out of 30; Economics and Discoverability are scored out of 20.
02Legibility
23.0 / 30
03Economics
19.5 / 20
04Discoverability
13.0 / 20
Highest-impact fix
Estimated gain +30 pointsAdd explicit identity and permission preflight tools
Expose machine-readable principal/tenant confirmation and a non-mutating permission check so agents can verify both before destructive actions.
Description evidence
Defects and rewrites.
17 defects found across the exposed tool descriptions. Suggested rewrites make purpose, inputs, boundaries, and returns easier for an agent to understand.
| Tool | Defect types | Suggested rewrite |
|---|---|---|
verify_agent |
params_unexplained |
Run a full 6-signal trust verification on an AI agent or wallet address. Pass the agent or wallet's address (a 0x-prefixed 40-character Ethereum address) to verify. Returns composite trust score (0-100), individual signal scores, risk flags, and recommendations. Use this before transacting with or delegating to an unknown agent. |
compare_agents |
params_unexplained no_return_description |
Compare two AI agents side-by-side on trust scores and security signals. Pass each agent's 0x-prefixed Ethereum address as address1 and address2. Returns a side-by-side comparison of both agents' trust scores and security signals. Useful for choosing between competing agents that offer similar capabilities. |
batch_verify |
params_unexplained |
Verify multiple AI agent addresses in a single call (max 10). Pass an array of 1 to 10 agent addresses (each a 0x-prefixed 40-character Ethereum address). Returns trust scores for each address. Efficient for evaluating a list of candidate agents. |
get_score |
params_unexplained |
Get a cached trust score for an AI agent address. Pass the agent's 0x-prefixed Ethereum address. Faster than full verification — returns the last computed score for that address without re-running signals. Free endpoint, no payment required. |
get_badge_url |
params_unexplained |
Generate a trust badge URL for an AI agent. Pass the agent's 0x-prefixed Ethereum address and optionally a style ('flat', 'pill', or 'detailed') to control the badge appearance. Returns an embeddable SVG badge showing the agent's trust score. Useful for displaying trust status in UIs or documentation. |
attest_agent |
params_unexplained no_return_description |
Score an AI agent and write an on-chain EAS attestation recording the trust score permanently. Pass the agent's 0x-prefixed Ethereum address. Costs $5 via x402 or requires an operator key. Returns the resulting on-chain attestation record (e.g., transaction details). Use when you need a verifiable, immutable trust record for an agent. |
get_attestations |
params_unexplained |
List all on-chain trust attestations written by AgentRadar. Optionally pass a page number (starting at 1) and a limit of up to 100 records per page to control pagination. Returns paginated attestation records including addresses, scores, and transaction hashes. Requires operator key. |
get_attestation_by_tx |
params_unexplained |
Look up a specific attestation by its transaction hash. Pass the hash as txHash (a 0x-prefixed 64-character hexadecimal string). Returns the full attestation record including the scored address, trust score, and on-chain details. |
lookup_identity |
params_unexplained |
Look up an AI agent's on-chain identity (ERC-8004 registration, ENS name, contract metadata). Pass the agent's 0x-prefixed Ethereum address. Returns who the agent claims to be and whether those claims are verified. |
lookup_reputation |
params_unexplained no_return_description |
Look up an AI agent's on-chain reputation given the agent's 0x-prefixed Ethereum address. Returns ERC-8004 feedback history, attestation count, and community signals. Shows how established and trusted the agent is in the ecosystem. |
resolve_agent_uri |
params_unexplained no_return_description |
Resolve an agent address to its registered URI, metadata, and service endpoints. Pass the agent's 0x-prefixed Ethereum address. Returns the registered URI, metadata, and service endpoints. Useful for discovering how to interact with a verified agent. |
check_scam |
params_unexplained |
Check if a wallet address is in the AgentRadar scam database. Pass the wallet's 0x-prefixed Ethereum address. Returns matching scam records with severity, chain, source, and notes. Use this as a quick safety check before any transaction. |
add_scam_wallet |
params_unexplained no_return_description |
Report a malicious wallet address to the AgentRadar scam database. Requires operator key. Pass the wallet's 0x Ethereum address, the chain on which it was observed, a severity level, and the source of the intel; notes is optional. Returns confirmation of the submission. Use when you need to flag known malicious activity. |
explain_score |
params_unexplained |
Get a human-readable explanation of an agent's trust score. Pass the agent's 0x-prefixed Ethereum address. Returns a narrative breakdown of each signal's contribution, risk factors, and actionable recommendations. Best for presenting trust analysis to end users. |
get_stats |
no_return_description |
Get AgentRadar dashboard statistics. Returns total verifications, attestations written, scam wallets tracked, and API usage metrics. Requires operator key. |
list_signals |
no_return_description |
List all trust scoring signals used by AgentRadar. Returns each signal with its weight and description. Helps understand how the composite trust score is calculated. |
get_pricing |
no_return_description |
Get AgentRadar API pricing information. Returns costs for each endpoint including x402 micropayment amounts and free tier details. |
Selection evidence
Confusable tool pairs.
6 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
get_attestations |
get_attestation_by_tx |
medium | Both retrieve on-chain attestation records with nearly identical names. A user asking to 'fetch the attestation for this transaction' or 'get attestation details' may route to get_attestations (list+pagination) instead of get_attestation_by_tx, since the tx-specific lookup is only obvious when the user explicitly mentions a transaction hash. |
get_score |
explain_score |
medium | Both take a single address and return trust-score information. A task like 'tell me about this agent's score' or 'why is the score 70?' sits between a cached score lookup and a narrative breakdown; identical input schemas give the agent no parameter-level hint to disambiguate. |
lookup_identity |
lookup_reputation |
medium | Both are single-address ERC-8004 on-chain lookups with overlapping vocabulary. Colloquial phrasing such as 'look up this agent on-chain' or 'what is this agent's reputation?' could plausibly select the wrong member, since identity claims and reputation/history are conceptually adjacent in normal language. |
get_attestations |
get_stats |
medium | A user asking 'how many attestations have been written?' or 'show attestation activity' maps to both: get_stats returns counts of attestations written, while get_attestations returns the underlying records. An agent may pick the list endpoint when the user actually wants dashboard totals. |
verify_agent |
attest_agent |
medium | Both compute and return a trust score for the same address-only schema; attest_agent additionally writes an on-chain EAS record for $5. A user task like 'verify this agent and record it' or 'verify and certify the trustworthiness of this agent' could route to verify_agent when on-chain attestation was requested, or to attest_agent when a simple verification was meant. |
verify_agent |
batch_verify |
low | Both are verify semantics for agent addresses. A user saying 'verify agent 0x...' might get routed to batch_verify (which accepts an array, including a single item) instead of the single-address verify_agent, though the list vs. single framing in the descriptions reduces this risk. |
Compare the field