Bodacc MCP Server

@pipeworx/bodacc

French official business-events bulletin (BODACC — Bulletin officiel des annonces civiles et commerciales): dated notices per company — creations, insolvency proceedings (redressement/liquidation judiciaire), business sales/transfers, deregistrations, modifications, and annual account filings.

Part of Pipeworx — an MCP gateway connecting AI agents to 1683+ live data sources.

Tools

• bodacc_company_events(siren | name, event_type?) — full BODACC event history for one company, newest first. siren is the same 9-digit id used by the entreprises-fr and insee packs — chain a SIREN looked up there straight into this tool (or the reverse).
• bodacc_search(q?, name?, event_type?, departement?, since?, until?, siren?) — open-ended search across all BODACC notices. q is whole-notice full text (name, activity, judgment/act text, address, ...); name is the separate, field-scoped form for matching the company name only. Also filters by event family, département, date range, and/or SIREN.
• bodacc_insolvency_feed(since, until?, departement?) — insolvency-only (familleavis=collective) notices published since a date, newest first. Built for polling/monitoring with an advancing since.
• bodacc_get_notice(id) — a single notice by its BODACC id.

Auth

Keyless. The upstream (DILA's Opendatasoft portal) requires no registration or API key for read access to this dataset.

Data sources

• https://bodacc-datadila.opendatasoft.com/api/explore/v2.1/catalog/datasets/annonces-commerciales/records — the live BODACC records API (Opendatasoft Explore v2.1). ~50M+ rows, updated as new bulletins publish (daily, business days).

Traps:

• offset + limit is capped at 10,000 for this dataset (confirmed live: offset=9999&limit=2 → 400 InvalidRESTParameterError). total_count in the API response is NOT capped — only how far pagination can reach into it is. Every tool here reports total_count, count (rows actually returned), and a truncated flag (offset + limit >= 10000 while total_count exceeds it) so a caller never reads the cap as the true count. Narrow with event_type / departement / a date range to stay under the cap on a broad query.
• registre is an array, not a scalar — ["805243102", "805 243 102"] (digits-only and space-grouped forms of the same SIREN). ODSQL registre="805243102" matches against the array directly; this pack always strips spaces from an input SIREN before building that filter, and reads registre[0] (digits-only) back out when shaping a record.
• Most upstream fields are per-family, not universal. jugement is populated for insolvency/conciliation notices, acte for creations/sales, radiationaurcs for deregistrations, modificationsgenerales for modifications, depot for account filings — the others are null on a given row. Each is itself a JSON-encoded string, not a nested object; this pack parses whichever of them are non-null into a single detail object per notice rather than exposing raw upstream field names.
• familleavis codes (pass to event_type): creation, immatriculation, collective (insolvency), conciliation, retablissement_professionnel, modification, radiation, dpc (account filings), vente (sale/transfer), divers. Source of truth: the ac_mapping_famille_avis dataset on the same Opendatasoft domain. event_type also accepts English aliases (sale/transfer→vente, insolvency/bankruptcy→collective, deregistration/deletion→ radiation, accounts→dpc, registration→immatriculation) — an unrecognized value is a thrown error naming the valid codes, never a silent total_count: 0 (fixed 2026-09-23: event_type: "sale" used to build familleavis="sale", an upstream field value that matches nothing).
• ODSQL's search() is field-scoped unless you omit the field. search(commercant, "...") matches the commercant field only (company name as filed); the unscoped search("...") matches every indexed field (activity description, judgment/act text, address, etc). bodacc_search's q uses the unscoped form on purpose — the field-scoped form silently returned total_count: 0 for q: "fonds de commerce" (a term that never appears in a company name) while the unscoped form returns 3,188 hits for the same query, confirmed live 2026-09-23. Use name instead of q to get the field-scoped, name-only match deliberately.
• Every call here hits the upstream endpoint directly, per request: the API is keyless, has no meaningful rate limit for normal query volumes, and a since-bounded monitoring query naturally stays well under the 10k pagination cap.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "bodacc": {
      "url": "https://gateway.pipeworx.io/bodacc/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/bodacc/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1683+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/bodacc_company_events \
  -H 'Content-Type: application/json' \
  -d '{"siren":"805243102"}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/bodacc_company_events. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "bodacc": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-bodacc"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-bodacc

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Bodacc data" })

The gateway picks the right tool and fills the arguments automatically.

More

• Docs and guides
• pipeworx.io

License

MIT