Bankruptcy Observer — API & MCP

Public documentation and registry metadata for the Bankruptcy Observer API and MCP server. This repository contains no application source code.

What this is

• Bankruptcy REST API — Programmatic access to US business bankruptcy data (search, case by EIN, docket items, cases by NAICS/state, and more).
• Bankruptcy MCP server — The same data and queries via the Model Context Protocol for AI assistants and LLM workflows.

Data covers US business bankruptcies (federal courts) only. Consumer-only filings and non-US jurisdictions are not included.

---

Free lookups (no token, no API key, no payment)

AI assistants and users can look up bankruptcy cases with no account, no API key, and no API token:

Free query	MCP tool	Example
Debtor name	search_bankruptcy_cases_tool	search_term: "JOY-CPW, INC." — no court required
Case number	get_case_by_case_number_tool	short_case_number: "26-10543" — no court required

• Call these tools without an API token. Only case number or name is needed; court is not required. If more than one case matches (e.g. same case number in multiple courts), the server returns all matching cases. Each result includes limited case info; full case details are available with a paid plan.
• All other queries (wildcard search, EIN, industry, state, date range, docket, etc.) require a subscription and authentication with either an API token or an OAuth Bearer access token.

Docs: https://mcp.bankruptcyobserver.com/docs or https://api.bankruptcyobserver.com/docs (GET, no auth). The MCP root GET https://mcp.bankruptcyobserver.com/ also returns a free_lookups hint.

---

Sign-up and access

You can sign up for a subscription through the MCP server: use list_plans_tool to see plans, then purchase_plan_tool with plan_id (e.g. professional, business, enterprise) to get a Stripe Checkout link. After payment you receive an API token; restricted queries can be authenticated with that token or with OAuth Bearer access tokens from the MCP OAuth flow. No need to go to the website to subscribe.

For questions about access, billing, or data coverage, use only the contact form at https://www.bankruptcyobserver.com/contact. Do not publish or use direct email addresses for contact.

Paid plans (Stripe via MCP tools)

• Professional — $1,500/month, up to 500 billable tool calls/month.
• Business — $3,500/month, up to 10,000 billable tool calls/month.
• Enterprise — $5,000/month, up to 150,000 billable tool calls/month (intended for multiple clients/teams).

Billable tool calls correspond to data tools invoked via MCP (for example, search, case by EIN, docket items, NAICS/state filters) or their REST API equivalents.

Authenticate using either an API token or an OAuth access token, sent in one of these ways:

• Authorization: Bearer <your-token>
• X-API-Key: <your-token>
• Api-Key: <your-token>

---

Connecting via OpenAI

Use OpenAI's Responses API to call this MCP server directly.

Free tier (no token):

from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="gpt-4.1",
    tools=[{
        "type": "mcp",
        "server_label": "bankruptcy_observer",
        "server_url": "https://mcp.bankruptcyobserver.com/mcp",
        "require_approval": "never",
    }],
    input="Search for bankruptcy cases for Acme Corp",
)
print(response.output_text)

Paid tier (with token):

response = client.responses.create(
    model="gpt-4.1",
    tools=[{
        "type": "mcp",
        "server_label": "bankruptcy_observer",
        "server_url": "https://mcp.bankruptcyobserver.com/mcp",
        "headers": {"Authorization": "Bearer YOUR_TOKEN"},
        "require_approval": "never",
    }],
    input="Get docket entries for case 24-10543",
)

Note: Do not set Content-Type or Accept headers manually — the OpenAI client sets these correctly for the MCP protocol. Overriding them will cause the server to return an HTML error page instead of a JSON response.

OAuth option: Connectors such as ChatGPT can also authenticate with MCP OAuth and send Bearer access tokens automatically after login. API token headers are still supported.

---

Connecting in ChatGPT (OAuth)

To connect this MCP server in ChatGPT with OAuth, create a new app/connector and enter the OAuth endpoints below exactly.

If the Registration URL is omitted, Dynamic Client Registration will fail and ChatGPT will connect without usable tool authorization.

OAuth endpoints

• Auth URL: https://mcp.bankruptcyobserver.com/authorize
• Token URL: https://mcp.bankruptcyobserver.com/token
• Registration URL: https://mcp.bankruptcyobserver.com/register
• Authorization server base: https://mcp.bankruptcyobserver.com/
• Resource: https://mcp.bankruptcyobserver.com/mcp

After connecting, ChatGPT should complete OAuth in-browser and then call tools using OAuth Bearer tokens. Legacy API-token headers remain supported for clients that use header auth.

---

Endpoints

Service	Endpoint	Auth
REST API	https://api.bankruptcyobserver.com/api/v1/...	Free lookups: no auth. Restricted queries: API token or OAuth Bearer token
MCP	https://mcp.bankruptcyobserver.com/mcp	Free lookups: no auth. Restricted queries: API token or OAuth Bearer token
Docs	https://api.bankruptcyobserver.com/docs or https://mcp.bankruptcyobserver.com/docs	None

The /mcp path is the MCP protocol endpoint (POST only). For human-readable documentation, use the Docs URLs above (GET, no auth).

---

Available queries and responses

The same queries are available on both the REST API and the MCP server. Each query returns JSON; case-level responses include the standard fields below unless noted.

Query list

#	Query	Inputs	What you get	Notes
1	Search	search_term, limit, skip, optional court_id or court_state	List of cases with standard fields	FREE (no key): exact name returns limited info; full details are available with a paid plan. Otherwise: prefix (starts-with) on name; use *term for contains. For case number use Case by case number.
2	Case by EIN	ein	One case (or none) + einInDatabase: "Yes" or "No"	Single match; flag says whether we have this EIN. (Restricted: API token or OAuth Bearer token.)
3	Docket items	shortCaseNumber, optional court_id or court_state, limit	Case identifier, dateDocketUpdated, list of entries (itemNumber, itemText, itemDate, …)	Case resolved by short case number + optional court. (Restricted: API token or OAuth Bearer token.)
4	Case by case number	shortCaseNumber; optional court_state or court_id for paid disambiguation	List (free) or one case (paid)	FREE (no key): case number only, no court required; returns all matching cases. For paid requests, you may provide court_state/court_id to select a specific case when multiple courts match.
5	Cases by NAICS	naics (2–4 digits), limit, optional start_date, end_date	List of cases with standard fields	NAICS must be 2–4 digits; more than 4 is rejected.
6	Cases by state	state (2-letter), limit, optional start_date, end_date	List of cases with standard fields	Optional date range on dateFiled.
7	Case status / dates	—	Returned as part of case data	dateFiled, dateClosed, dateDismissed; isOpen / isClosed / isDismissed.

Pagination: List responses support limit and skip where applicable; responses include total or similar where useful.

Date range: For NAICS and state queries, optional start_date and end_date filter on dateFiled. Search does not support date filters.

Standard fields (case-level)

Returned for any query that returns case rows:

Field	Description
name	Debtor/case name
court	Court name
shortCaseNumber	Short case number (e.g. 22-12345)
dateDocketUpdated	When docket/case data was last updated in our system
dateFiled	Filing date

Optional (included when relevant): courtState, chapter, NAICS, industry, dateClosed, dateDismissed, isOpen / isClosed / isDismissed, assetAmount, liabAmount.

EIN query: einInDatabase

For Case by EIN the response always includes:

• einInDatabase — "Yes" if we have at least one case with this EIN (and the case payload); "No" if we have no case with this EIN (case payload null or omitted).

Docket response shape

For Docket items you get:

• Case identifier (e.g. shortCaseNumber, court).
• dateDocketUpdated — when this case’s docket was last updated in our system.
• entries — array of objects with at least itemNumber, itemText, itemDate (and optionally pageCount, pacerDocId).

---

API vs MCP mapping

Query	REST API (POST)	MCP tool
Search	/api/v1/search	search_bankruptcy_cases_tool
Case by EIN	/api/v1/case/ein	get_case_by_ein_tool
Docket items	/api/v1/docket	get_docket_entries_tool
Case by case number	/api/v1/case/by-number	get_case_by_case_number_tool
Cases by NAICS	/api/v1/cases/by-naics	get_cases_by_naics_tool
Cases by state	/api/v1/cases/by-state	get_cases_by_state_tool

Same backend for both; response shape is identical.

---

Full documentation (canonical)

The full spec (same content as above, plus any updates) is also served with no auth at:

• API docs: https://api.bankruptcyobserver.com/docs
• MCP docs: https://mcp.bankruptcyobserver.com/docs

Both URLs return the same Markdown. Use them for the latest version or for AI/bot discovery.

---

Registry

This repo includes a server.json for the Official MCP Registry. It describes the remote MCP server and how to connect (URL + API token header). Use it when publishing to the registry via the Registry CLI; the registry points to this GitHub repo for metadata only.

---

License

Documentation and metadata in this repository are provided for discovery and integration. The Bankruptcy Observer product, API, and MCP service are offered under separate terms; see the product website.