canonical: https://jentic.com/apis/aidecisions.ai/aidecisions # AI Decisions Compliance API The AI Decisions Compliance API screens blockchain addresses and transactions for compliance risk, returning sanctions hits, mixer exposure, AI-agent detection, and a population-calibrated risk tier across Ethereum, Bitcoin, Tron, Base, Arbitrum, and Gnosis. Each label resolves to a primary public source so a finding can be traced. It also scores and finds behaviourally similar addresses, investigates an address's history on served chains, and exports investigation reports. A separate lookup checks UK companies in the Companies House register by name or number. ## For AI agents Screen blockchain addresses and transactions for sanctions, mixer exposure, AI-agent patterns, and a risk tier across six chains for AI agents, with source-backed labels. ## Scope Does not freeze funds, file suspicious-activity reports, or provide investment advice. Use for screening blockchain addresses and transactions for compliance risk only. ## Capabilities - Screen a blockchain address for sanctions hits, mixer exposure, and a risk tier - Score a transaction's compliance risk and detect AI-agent transaction patterns - Investigate an address's history across served chains and export the report - Label an address and find behaviourally similar addresses - Look up a UK company in the Companies House register by name or number - List compliance check history and the available tier definitions ## Use cases ### AI compliance screening agent An AI agent in an onboarding flow screens a counterparty wallet before approving it. Through Jentic the agent searches for 'screen a blockchain address', loads the matching AI Decisions operation with its input schema, and screens the address with the key injected at call time. The response carries sanctions and mixer exposure plus a risk tier, each tied to a primary source the agent can cite. Example prompt: Call POST `/v1/screen/{chain}/{address}` for a wallet and flag it when the risk tier exceeds a threshold ### Transaction monitoring A monitoring system scores transactions for compliance risk and watches for automated behaviour. AI Decisions scores a transaction and separately detects AI-agent transaction patterns, so a monitor can combine both signals. History and export endpoints let a team keep an auditable record of what was checked. Example prompt: Submit a transaction to POST `/compliance/transaction-check` and run POST `/compliance/agent-detection` on the same address ### Address investigation and clustering An investigator wants an address's history and others that behave like it. AI Decisions investigates an address across served chains, labels it, and finds similar addresses by behaviour. An agent can pull the investigation, read the labels, and expand to similar addresses to map a cluster. Example prompt: Investigate an address with GET `/investigate/{chain}/address/{address}`, read labels with GET `/v1/ai-agents/labels/{address}`, then find similar ones ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/v1/screen/{chain}/{address}` | Screen an address for compliance risk | | GET | `/investigate/{chain}/address/{address}` | Investigate a blockchain address | | POST | `/compliance/transaction-check` | Score transaction compliance risk | | POST | `/compliance/agent-detection` | Detect AI-agent transaction patterns | | GET | `/v1/ai-agents/labels/{address}` | Get labels for an address | | GET | `/compliance/history` | List compliance check history | | GET | `/investigate/chains` | List served chains | | GET | `/tiers` | Get tier definitions | ## Key resources - **Screening** — Address screening and transaction compliance risk scoring with AI-agent detection - **Investigation** — Address investigation across served chains, labels, similar-address search, and report export - **Companies** — UK Companies House register lookups by name or number - **Account** — Compliance check history and tier definitions ## AI readiness This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use. - **Score:** 59 / 100 - **Maturity:** Foundational - **Dimensions:** - Foundational Compliance: 93 / 100 - Developer Experience & Jentic Compatibility: 62 / 100 - AI-Readiness & Agent Experience: 41 / 100 - Agent Usability: 94 / 100 - Security: 50 / 100 - AI Discoverability: 53 / 100 - **View full report:** https://jentic.com/apis/aidecisions.ai/aidecisions/scorecard - **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/ - **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview ### Score it yourself Every API in the directory is allowlisted, so you can re-score it with no key required. - **Score your own API:** https://jentic.com/scorecard.md - **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md ```sh npx @jentic/api-scorecard-cli score ``` ## Why Jentic - **Setup:** Wiring AI Decisions by hand means obtaining a tenant key, sending it in the X-API-Key header, and threading the chain and address through the screening and investigation paths. Through Jentic you install once, import AI Decisions from the API Directory, store the key once, and your agent calls it. - **Permission scoping:** The chain and address sit in the URL path such as `/v1/screen/{chain}/{address}`, so a rule can pin your agent to screening and investigation for the chains it needs. You choose which operations it may call, so report export and the Companies House lookups are not included unless you add them. - **Credential handling:** Your AI Decisions tenant key is stored once, encrypted, by your own Jentic One instance and attached as the X-API-Key header at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'screen a blockchain address' or 'score a transaction', and Jentic returns the matching AI Decisions operation with its input schema, including the chain and address, so the agent calls the right endpoint. ## Related APIs - **Onfido Public API** — Identity verification and KYC checks for people rather than on-chain address screening - **Bitquery GraphQL API** — On-chain data and transaction graphs across many blockchains - **Moralis EVM API** — Wallet balances, token transfers, and EVM chain data for enrichment ## FAQ ### What authentication does the AI Decisions Compliance API use? It authenticates with a tenant API key sent in the X-API-Key request header, as declared in its OpenAPI spec, with GET /health left open as a liveness check. Through Jentic the key is held by your own instance and attached per call, so an agent screening addresses never sees it. ### Which chains can the AI Decisions API screen? The screening and investigation endpoints cover Ethereum, Bitcoin, Tron, Base, Arbitrum, and Gnosis, with the chain supplied in the request path. GET `/investigate/chains` reports the served chains and whether a transaction history is available for each. ### What are the rate limits for the AI Decisions Compliance API? Limits are tied to the tier on your key rather than a single fixed ceiling in the spec, and GET /tiers returns the tier definitions. Check the AI Decisions documentation for the per-tier request allowance before running bulk screening jobs. ### Is there an AI Decisions MCP server? You do not need an MCP server to give your agent AI Decisions. Jentic connects it directly from the API Directory: import it, store your tenant key once, and your agent calls the screening and investigation operations it needs on demand. ### How do I screen a wallet address through Jentic? Search Jentic for 'screen a blockchain address', load the POST `/v1/screen/{chain}/{address}` operation, and run it with the chain and address. The response carries the risk tier and source-backed labels, with the key attached by your instance. To run it on your own infrastructure, install Jentic One from its GitHub repo. ### Can I limit what my agent is allowed to do with the AI Decisions API? Yes. The chain and address travel in the URL path, so a rule can allow the agent to screen and investigate addresses while leaving report export and the Companies House lookups out unless you add them. The agent stays bound to the operations you grant, and every call is logged.