Integrations

conmcp is the plan-reading front end of the ContractorTakeoff suite — a set of small, local-first tools for taking a project from plan set to bid. conmcp stops at quantities on purpose: pricing and labor are downstream tools' jobs.

What's real today vs. roadmap

Honest status: the JSON and CSV exports exist today and are designed to feed the tools below. The deeper integrations (evidence hooks, automated handoffs) are roadmap, not shipped. This page says which is which.

contractor-bid — JSON export (works today)

generate_takeoff_report with fmt: "json" produces the contractor-bid takeoff format: a line_items array in the BOM shape (id/section/item/description/qty/uom/ status/source/notes/customFields), plus project and extraction metadata. See the exact shape in the takeoff schema.

The field names are a deliberate contract — conmcp's schema was designed around contractor-bid's format, not translated to it after the fact. The workflow:

# LLM builds the takeoff via MCP, saves JSON…
# or render an existing TakeoffReport from the CLI:
conmcp report oakridge-takeoff.json --fmt json --out oakridge-bid-import.json

…and the resulting file drops into contractor-bid as takeoff line items, with confidence and needs_review riding along in customFields so the review state survives the handoff.

pma-standalone — CSV export (works today)

fmt: "csv" renders the takeoff with columns matching the pma-standalone takeoff export layout:

Section, Item, Description, Qty, UOM, Status, Source, Location,
Confidence, NeedsReview, Assumptions, CustomFields

It also opens clean in Excel, Google Sheets, or anything else that eats CSV — column order and formatting are documented in the takeoff schema.

contractor-bid-pro — Finding/evidence hooks (roadmap)

Planned, not shipped: mapping conmcp output onto contractor-bid-pro's Finding model, so each takeoff item arrives with its evidence attached — the sheet reference, the rendered page crop it was read from, and the extraction confidence. The schema already carries the raw material (source, confidence, custom_fields); the wiring is future work. Track it on the roadmap.

Everything else

The exports are plain JSON, CSV, and Markdown — nothing proprietary. If you're building your own estimating pipeline, the takeoff schema is versioned (schema_version: 1) and the field names won't change casually; that's a stated maintenance rule of the project.