SecureLend Underwriting Agents
SecureLend — ИИ-агенты для андеррайтинга сделок венчурных фондов через ChatGPT и Claude
От сообщества: Добавлен пользователем или импортирован; проверьте владельца перед подключениемРаботаетБез входаГлобальныйБесплатноТолько чтение
Что умеет
- Проверять питч-деки по инвестиционному рубрику
- Извлекать структурированные финансовые данные из документов
- Проверять сущности на комплаенс и публичные подачи
- Собирать количественный анализ и оценку рисков
- Формировать меморандум для инвестиционного комитета
Какие данные видит
Платежи проходят через отдельный мандат расходов с лимитами; ассистент не видит номера карт, только замаскированные сведения.
Нужен ли аккаунт
Не нужен: сервер работает без входа
SecureLend помогает фондам проверять питч-деки против инвестиционного рубрика, извлекать финансовые данные, проверять контрагентов и собирать меморандум для инвестиционного комитета в общем рабочем пространстве по сделке. Платные операции авторизуются и оплачиваются через delegare.dev.
Как подключить
Как подключить SecureLend Underwriting Agents к Claude
Адрес сервера, который нужно вставить в настройках Claude:
https://securelend-agents.run.tools
- Откройте Claude (сайт claude.ai или приложение для компьютера).
- Перейдите в Настройки → Коннекторы (в английском интерфейсе: Settings → Connectors).
- Нажмите Добавить пользовательский коннектор (Add custom connector).
- Вставьте адрес сервера, скопированный ниже, в поле URL удалённого MCP-сервера и нажмите Добавить.
- В чате нажмите + → Коннекторы и включите новый коннектор.
Пользовательские коннекторы доступны на тарифах Free, Pro, Max, Team и Enterprise. На Free можно добавить только один. В Team и Enterprise коннектор сначала добавляет владелец организации в Настройки организации → Коннекторы.
Авторизация
После нажатия Добавить откроется окно входа в сервис. Войдите под своим аккаунтом и подтвердите доступ. Пароль Claude не видит.
Как подключить SecureLend Underwriting Agents к ChatGPT
Адрес сервера, который нужно вставить в настройках ChatGPT:
https://securelend-agents.run.tools
- Откройте ChatGPT в браузере (chatgpt.com). Нужен тариф Plus, Pro, Business, Enterprise или Edu.
- Один раз включите режим разработчика: Настройки → Приложения → Дополнительные настройки → Режим разработчика (Settings → Apps → Advanced settings → Developer mode).
- Откройте Настройки → Коннекторы и нажмите Создать (Create).
- Заполните форму: Название (любое), Описание (одна фраза, что делает сервис), URL MCP-сервера (скопируйте ниже).
- В поле Аутентификация выберите OAuth, если сервис требует вход, иначе Нет. Нажмите Создать.
- В новом чате откройте + → Приложения/Коннекторы и включите коннектор.
Название раздела у OpenAI менялось (Connectors → Apps/Plugins). Если пункт называется иначе, ищите «режим разработчика» в настройках. В рабочих пространствах Business/Enterprise администратор должен разрешить пользовательские коннекторы.
Авторизация
При первом использовании ChatGPT откроет окно входа в сервис. Войдите и подтвердите доступ.
Как подключить SecureLend Underwriting Agents к Cursor
Адрес сервера, который нужно вставить в настройках Cursor:
https://securelend-agents.run.tools
Быстрый способ: нажмите кнопку Открыть в Cursor ниже, Cursor сам предложит добавить сервер.
Вручную:
- В Cursor откройте Settings → Cursor Settings → MCP и нажмите Add new global MCP server.
- Вставьте JSON, скопированный ниже, в открывшийся файл
~/.cursor/mcp.json(для одного проекта:.cursor/mcp.jsonв корне проекта). - Сохраните файл. Сервер появится в списке MCP; при необходимости пройдите авторизацию по кнопке рядом с ним.
Авторизация
Если сервис требует вход, рядом с сервером в списке MCP появится кнопка авторизации.
Список инструментов сервера (50)
Технические названия из tools/list. Нужны только разработчикам.
| setup_spending_mandate | Initiate the one-time browser setup flow so the user can connect their payment method and set spending limits. Returns a URL the user must visit. Returns sessionToken for polling. |
| poll_setup_session | Check whether the user has completed the payment setup flow. Call this after presenting the setup URL. Returns the intentMandate once complete — store it in agent context for future payments. |
| check_mandate_balance | Check remaining monthly budget and masked payment methods for a spending mandate. Never returns card numbers or wallet seeds — only masked summaries. |
| authorize_agent_payment | Execute a payment through the Delegare vault using AP2. The agent presents its Intent Mandate (SD-JWT-VC). Spending limits are enforced server-side. IMPORTANT: amountCents is in US cents — divide by 100 for the dollar amount (e.g. amountCents=50 means $0.50, NOT 50 dollars or 50 USDC). |
| delegare_fetch | Fetch a URL. If the resource requires payment via x402, this tool will automatically use the provided spending mandate to authorize the payment and retrieve the data. Supports both GET and POST. |
| revoke_mandate | Immediately revoke a spending mandate. After revocation, no further charges can be made with this intentMandate. The user can create a new mandate at any time. |
| link_to_firm | Roll out SecureLend at your firm. Creates a firm workspace, makes you admin, and switches your free tier to a firm-shared bucket so teammates can use the same agents under one budget. Use this when an individual user is ready to bring SecureLend to their bank, VC firm, insurer, or PE shop. |
| create_deal_workspace | Create a new deal workspace to group multiple documents and analysis tasks. Returns a Workspace ID. |
| get_deal_workspace | Retrieve all tasks, documents and analysis associated with a workspace. |
| submit_documents | Reserve a pre-signed upload URL so the user can upload a document directly to SecureLend's secure storage from the browser. Returns a documentId that subsequent agents (intelligence, extraction, risk discovery, etc.) can reference. |
| display_upload_documents_form | Display an interactive form for the user to upload one or more documents to SecureLend's secure storage. Use this to collect documents before running classification, extraction, or risk-discovery agents. |
| run_document_intelligence | Instant, accurate document sorting. Automatically identifies and categorizes files for any underwriting domain. Cost: $0.06 per page. |
| pitch_deck_precheck | Score an inbound pitch deck against your firm's rubric. Returns overall fit %, per-criterion breakdown, and follow-up questions an analyst should ask before spending time on a call. PDF only. Cost: $0.50 per check (first 50/month free). |
| run_data_extraction | Structured data extraction from unstructured files. Uses domain-specific blueprints. Cost: $0.15/page (Std) | $0.25/page (Custom). |
| run_quantitative_analysis | Performs math based on Deal Type (e.g., LTV, DSCR, Loss Ratios, Burn Multiples). Cost: $1.50 per risk score. |
| run_risk_discovery | Detects anomalies, tampering, and inconsistencies in deal data. Cost: $0.24 per page. |
| run_professional_memo | Generates a professional underwriting analysis memo — credit, equity investment, or insurance — using the SecureLend AI pipeline. Supports three source modes: • workspace (default when workspaceId is provided): synthesises all prior agent outputs into a full memo. Two pre-built workflows: – Equity/VC: upload deck → pitch_deck_precheck → run_quantitative_analysis → run_professional_memo(workspaceId) → Investment Committee memo – Insurance: upload submission → run_document_intelligence → run_entity_compliance → run_risk_discovery → run_professional_memo(workspaceId) → Underwriting memo • lending_application: traditional credit memo from a loan application (sourceId = applicationId). • document: memo from a single processed document (sourceId = documentId). Cost: $4.99 per memo. |
| get_memo_status | Check the status of a memo generation job started by run_professional_memo. Call this ~75 seconds after starting a memo, or whenever the user asks to check progress. Returns the full memo sections when COMPLETED so the widget can render them inline. |
| run_entity_compliance | Real-time AML, KYC, KYB, sanctions, and PEP screening via ComplyAdvantage (default) or Refinitiv World-Check. Results are stored in the workspace so the underwriting memo pipeline can reference compliance status. Cost: $0.60 per check. |
| fetch_public_filings | Fetch publicly available company information and SEC filings from EDGAR — free, no API key required. Supports lookup by ticker symbol, company name, or CIK. Returns company metadata, SIC classification, and links to recent 10-K / 10-Q / 20-F filings. Use this for equity analysis before running pitch_deck_precheck or generating an investment memo — it provides audited financial context and business description that grounds the AI analysis. Results are stored in the workspace. Cost: Free. |
| fetch_docsend_document | Fetch a pitch deck or dataroom document directly from a DocSend share link — no manual downloading and re-uploading required. The document is ingested into SecureLend and returned as a documentId ready for pitch_deck_precheck, run_document_intelligence, and the investment memo pipeline. Also returns page-level engagement analytics (time spent per slide, viewers) which provide useful signal about founder presentation strategy. Requires: configure_data_provider(provider='docsend', apiKey='...') run once first. Note: DocSend API access requires a Business or Enterprise plan — if you're on a lower tier, upload the deck manually via display_upload_documents_form instead. Typical VC workflow: fetch_docsend_document(link, workspaceId) → pitch_deck_precheck(documentId, workspaceId) → run_professional_memo(workspaceId) Cost: Free (infrastructure only — you pay DocSend directly via your contract). |
| fetch_crm_context | Pull deal context from Affinity CRM — the relationship intelligence layer most VC firms use. Returns company profile, custom field values (stage, conviction score, source, sector tags etc.), recent interaction notes, key contacts, and linked opportunities. Storing this in the workspace grounds the investment memo in your firm's prior relationship history and pipeline context — the AI knows your conviction level, who made the intro, and what questions your team has already raised. Requires: configure_data_provider(provider='affinity', apiKey='...') run once first. Get your API key from Affinity → Settings → API Keys. Typical workflow: fetch_crm_context(companyName, workspaceId) ← pull firm's prior context fetch_docsend_document(link, workspaceId) ← or manual upload pitch_deck_precheck(documentId, workspaceId) run_professional_memo(workspaceId) ← memo references all context Cost: Free (infrastructure only). |
| fetch_company_intelligence | Pull company intelligence from Harmonic — funding history, headcount growth, traction signals, and market context. Use this before a pitch deck arrives to pre-populate the workspace with public signals: how much has the company raised, who invested, is the team growing or shrinking, what stage are they at. Lookup by domain (most reliable) or company name. The investment memo pipeline references workspaceData.fetch_company_intelligence for the Financial Analysis and Company Overview sections. Requires: configure_data_provider(provider='harmonic', apiKey='...'). Get API key at console.harmonic.ai. Cost: Free (infrastructure only). |
| fetch_portfolio_metrics | Pull reported KPIs and investor updates from Visible.vc for a portfolio company. Returns the latest metric values (ARR, MRR, headcount, runway, burn rate — whatever the company has set up in Visible) and recent investor update notes. Use case: follow-on investment decisions, portfolio reviews, or LP reporting. The investment memo pipeline references workspaceData.fetch_portfolio_metrics for the Financial Analysis section — actual reported numbers rather than estimates. Requires: configure_data_provider(provider='visible', apiKey='...'). Token is user-scoped — request from Visible Settings or email dev@visible.vc. The company must have a Visible profile and your token must have access to it. Cost: Free (infrastructure only). |
| run_portfolio_analytics | Portfolio and pipeline visibility in real time. Cost: Free. |
| submit_underwriting_case | Promotes a workspace into a formal case in the Master Ledger. Call this after all analysis is complete. |
| list_custom_blueprints | List all your custom document extraction blueprints. |
| create_custom_blueprint | Create a custom blueprint. Two kinds: 'extraction' (structured data extraction from docs) and 'rubric' (VC pitch-deck scoring rubric for pitch_deck_precheck). |
| configure_data_provider | Register your own API credentials for a compliance or data provider — ComplyAdvantage, Refinitiv World-Check, Experian, Dun & Bradstreet, etc. Once registered, all checks run directly against your provider contract; SecureLend only charges for infrastructure (orchestration calls above the free tier). Your credentials are AES-256 encrypted at rest and never logged. Use this once per provider — credentials persist until you rotate them. |
| run_deal_screening | Single-call screening pass. Returns proceed/pass/maybe + rationale. |
| run_dd_workflow | Multi-step DD orchestration (checklist -> meeting prep -> IC memo -> VCP). |
| list_rubric_templates | List the SecureLend system rubric catalog — pre-built scoring rubrics for common investment scenarios. Returns compact summaries; use create_custom_blueprint(kind='rubric') to clone and customise one. Optional filters: vertical ('vc','pe','insurer') and stage ('pre_seed','seed','series_a','series_b','growth'). |
| initiate_payment_setup | Start the Delegare payment setup flow so you can pay for SecureLend agent tasks beyond the free tier. Returns a setup URL — open it in a browser to authorize a spending budget via card or crypto wallet. You will need to complete one human step: enter card details, or sign one transaction in a crypto wallet. After that, all payments are automatic. Call check_payment_setup with the returned session_token to confirm completion and retrieve your mandate. |
| check_payment_setup | Check whether the Delegare payment setup flow has been completed. Call this after the user has visited the setup URL from initiate_payment_setup and authorized their budget. Returns 'complete' with the mandate token if ready, 'pending' if still in progress, or 'expired' if the session timed out. Once complete, your mandate is saved and all subsequent agent tasks will charge it automatically. |
| setup_fiat_mandate | Set up a Stripe card payment mandate via Delegare for API task billing. Use this when USDC payments are unavailable or you prefer to pay by card. Delegare hosts the secure card entry — SecureLend never sees card details. Spending limits: $5.00 per transaction, $50.00 per month. Once set up, all API tasks (pitch deck pre-checks, underwriting, etc.) are charged automatically. Returns a setup URL — the user visits it once to enter their card details. |
| get_attributed_fact | Look up the full details and provenance of an attributed fact by its ID. |
| run_comps_analysis | Build public comps table in Excel format. |
| run_dcf_valuation | Full DCF with WACC + sensitivity matrix. |
| run_competitive_landscape | Map positioning across the sector. |
| run_sector_overview | Industry landscape deep dive. |
| run_earnings_analysis | Quarterly beat/miss analysis from 10-Q/8-K. |
| generate_buyer_list | Potential acquirers universe mapping and ranking. |
| generate_teaser | Anonymous one-pager for sell-side processes. |
| generate_pitch_deck_content | Returns structured pitch book content for manual or automated assembly into PPTX. |
| build_data_pack | Normalizes financial data from filings and documents into an IC-ready format. |
| run_merger_model | Calculate accretion/dilution and pro forma financial impact of a transaction. |
| generate_cim_content | Generates draft sections for a Confidential Information Memorandum using workspace data. |
| generate_strip_profile | Generates content for a 1-4 slide company profile 'strip'. |
| track_deal_status | Update and query transaction milestones, deadlines, and pipeline status. |
| submit_tool_feedback | Record a user's feedback on an IB tool's output to improve future results. Call this whenever the user replies to a tool's '📊 How useful was this? Reply 1–9' prompt with a number (optionally with notes/edits). Set `toolName` to the tool they're rating, map their number to `rating` (1–9), put any free-text into `comment`, and if they suggested concrete changes put those in `suggestedEdits`. invocationId is optional — include it only if the tool's structuredContent surfaced one. |