> ## Documentation Index
> Fetch the complete documentation index at: https://docs.blockgence.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Tools

> The tools the MCP server gives your assistant, what each one does and what it costs.

Each tool calls an API endpoint with your key and uses the same credits as that endpoint. The free ones are useful to
get oriented before spending.

> [!NOTE]
>
> With every tool, your assistant gets the same rules as the API: a source failure is not «no data», a partial result
> carries `degraded_sources`, and a lead is presented as a lead, never as a fact.

## Coverage and account

| Tool | What it does | Credits |
| - | - | - |
| `list_chains` | The supported networks and what can be queried on each one. Worth using before claiming what can be seen on a network. | 0 |
| `source_health` | The status of the data sources: which are answering, which are down or out of plan, and until when. | 0 |
| `account_usage` | Your plan, your limits and what you have used this month. | 0 |

## Screening and attribution

| Tool | What it does | Credits |
| - | - | - |
| `screen_address` | An address's risk: category, OFAC sanctions with their history, score and recommendation (`block`, `review` or `allow`). Accepts EVM, Tron and Monero. | 3 |
| `get_address_labels` | The known labels of an address, including the ones you added, with its risk and sanctions status. | 1 |
| `enrich_address` | An address's on-chain data and its main labeled counterparties. | 4 |
| `check_contract_security` | A contract's security: verified, honeypot, taxes, proxy, holders and malicious flags. If the source fails, it says so: it never answers «not malicious» because of a failure. | 3 |
| `get_funding_source` | Who funded a wallet first, resolved against known entities. | 1 |
| `probe_chains` | Which networks an address has real activity on. A network that could not be queried shows up as unknown, not as «no activity». | 2 |
| `osint_web` | What is publicly known about an address on the open web. Only a result that shows the full address counts as a mention. | 5 |
| `resolve_ens` | Resolves ENS names (`.eth`). | 2 |

## Fund tracing

| Tool | What it does | Credits |
| - | - | - |
| `trace_funds` | Traces funds from an address: risk, exposure by category, counterparties, notable entities and largest flows. | 25 |
| `trace_from_tx` | Traces from a loss transaction: what left the victim and where it ended up (exchanges, bridges, mixers). | 25 |
| `follow_bridge` | From a bridge deposit, rebuilds the other side of the crossing (destination network and recipient) using the protocol's identifier. | 25 |
| `unswap_transaction` | Follows value through a swap: which asset went in, which came out, how much and to whom. | 10 |
| `demix_tornado` | From whoever deposited into Tornado Cash, the withdrawals that probably match. They are leads with a confidence, not proof. | 15 |
| `get_transaction` | A transaction's internal movements. | 5 |
| `get_onchain_messages` | Text messages in an address's transactions. | 4 |

## Contracts, tokens and transactions

| Tool | What it does | Credits |
| - | - | - |
| `inspect_contract` | Who controls a contract and who can move, mint or freeze its funds, read on-chain at a fixed block. | 10 |
| `simulate_transaction` | Simulates a transaction before signing it: whether it goes through or reverts, balance changes, approvals it grants and delegations. | 15 |
| `token_profile` | A token's profile to analyze a rug pull: deployer, pools, liquidity, mints, sales and blacklist. | 25 |
| `decode_transaction` | A transaction's decoded execution: call tree, events, balance changes and flashloans. | 15 |

## Approvals, drains and victims

| Tool | What it does | Credits |
| - | - | - |
| `get_approvals` | Every approval an address gave and its status today (open, closed or expired), with how much each spender can move. | 10 |
| `approvals_by_spender` | Reverse lookup: who gave approval to a spender and which approvals are still open. | 15 |
| `signature_drains` | Outflows from an address in transactions it did not send, with the mechanism that authorized them (signature, permit, delegation). | 20 |
| `get_delegation` | An address's EIP-7702 delegations and the transaction that installed the current one. | 3 |
| `detect_address_poisoning` | Address poisoning: addresses similar to a victim's counterparties and the loss, if there was one. | 5 |
| `drainer_victims` | A drainer's victims, with what was stolen in USD and the approvals still open. | 25 |
| `scam_victims` | A scammer's victims with evidence, kept apart from indicators and from everyone else who sent them funds. | 40 |
| `classify_compromise_vector` | For each loss transaction, who authorized the outflow, with the evidence and what was ruled out. It describes the mechanism, never blame. | 30 |

## Incidents

| Tool | What it does | Credits |
| - | - | - |
| `incident_playbook` | From an exploit transaction, starts a job that decodes it, identifies the attacker, computes what was stolen, traces it and builds the case. | 40 plus each completed step |
| `stolen_funds_tracker` | Follows the stolen funds of a hack: how much was returned, frozen, reached exchanges, mixers or bridges, or is still sitting idle. | 40 to start or recompute; 0 to read |
| `incident_postmortem` | A draft post-mortem of a case's incident, with the on-chain part cited. | 40; 0 if it was already built in the last 24 h |
| `exploit_attempts` | The attacker's failed attempts, with the reason for each revert. | 10 |
| `fund_returns` | What was returned from a hack and how it is classified, against the SEAL Safe Harbor agreement or the offered bounty. | 40 |
| `safe_harbor_lookup` | Whether an address is covered by an on-chain SEAL Safe Harbor agreement. | 5 |
| `issuer_freeze_status` | Whether an address is frozen today by the USDT or USDC issuers, and when it was frozen or released. | 3 |
| `freeze_evidence_pack` | The package to request a freeze from the authority or the issuer, from a saved case. Blockgence sends nothing. | 30 |

## Monero

| Tool | What it does | Credits |
| - | - | - |
| `get_monero_transaction` | The public record of a Monero transaction. Monero does not publish senders, recipients or amounts: there is no tracing. | 5 |
| `verify_monero_payment` | Verifies an XMR payment with the payer's tx key or a signed proof, without the key leaving the server. | 10 |

> [!NOTE]
>
> The Monero tools need a Monero node configured on the server. Without one, they return `source_unavailable` with
> `reason: "not_configured"`.

## Guard

| Tool | What it does | Credits |
| - | - | - |
| `list_guard_alerts` | The active alerts of your watches and of the ones your organization shares. | 0 |
| `get_guard_alert` | One alert: the transaction's events, its history and the hops followed. | 0 |
| `create_guard_watch` | Watches an EVM address on a network, your own or an attacker's. It respects your plan's watch limit. | 0 |

The Guard tools need a key assigned to a person.

## Example prompts

* «Screen 0x098b716b8aaf21512996dc57eb0615e2383e2f96 and tell me whether I should block it.»
* «Is the contract 0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48 safe or is it a honeypot?»
* «Follow the money from 0x098b… three hops and tell me which exchanges or mixers it reached.»
* «How much have I used this month?»

> [!TIP]
>
> Labels you added with your key are visible to every tool. Labels you added inside a case are not, because the MCP does
> not work inside a case.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.