- Brand
- CompanyCam
- Category
- Operations
- Primary Subcategory
- Field Service Management Software
Integration details
Description
CompanyCam helps contractors find projects, review jobsite photos and documents, add comments and tasks, and keep customers, checklists, and estimates up to date through ChatGPT. It only accesses data available to your connected CompanyCam account.
- Integration type
- Plugin
- Verification status
- Not applicable
- Platform
- ChatGPT
- Primary Subcategory
- Field Service Management Software
- Secondary Subcategories
- None listed
- Brand
- CompanyCam
- Access
- Account required
- First tracked
- 2026-11-20
- Tool count
- 152
- Market
- US
The Primary Subcategory used for this profile’s headline score.
Other Subcategories where the Integration is listed.
Not discoverable in ChatGPT yet
ChatGPT now suggests Plugins on its own when they match a user's request, but not every Plugin is being surfaced yet.We haven’t seen CompanyCam suggested yet. Sign up and we’ll track it, and your Discoverability Score will appear here once ChatGPT starts surfacing it.
What discovery looks like

Get alerts for CompanyCam
Get updates when CompanyCam’s Discoverability Score or category rank changes.
Competing in ChatGPT Field Service Management Software
View Category152 tools agents can invoke
Add a single contact to a customer. Requires a `name` (or `first_name`). Returns the created contact. Fires customer.contact_created.
add_customer_contact
Add a project to a project group. Both must belong to the current company. Idempotent: re-adding a project already in the group succeeds. Returns the project.
add_project_to_group
[beta] Add a supplier item to the company price book by SKU, with the supplier's name and price. If the SKU is already in the price book, returns that item unchanged. Find SKUs with search_supplier_catalog. Needs manager access, and stays off until an admin or manager turns on price book updates through AI tools under Integrations → CompanyCam MCP. Beta — may change shape or behavior without the standard deprecation process.
add_supplier_catalog_item
Archive a project: hide it from the active list while keeping it fully intact. Reversible with unarchive_project. This is NOT deletion — to soft-delete a project, use delete_project. Fires project.archived.
archive_project
Assign a user to a project. Both must belong to the current company; the user must be active. The assigned user is emailed and push-notified about the assignment. If the user was previously unassigned from the project, that assignment is reactivated and its role reset to the user's current role. Returns the assigned user's profile (id, name, email, role), not an assignment record.
assign_user_to_project
Clear one custom field's value on a project, leaving the field unset. The company's field definition is untouched, and every other project keeps its own value. Clearing a field that was never set succeeds. Returns confirmation.
clear_project_custom_field
Post a discussion COMMENT on a photo as the authenticated user (appends; supports @mentions/notifications, and webhooks subscribed to comment.created receive the event). To set the photo's own caption, use add_photo_description.
comment_on_photo
Post a discussion COMMENT on a project as the authenticated user (appends; supports @mentions/notifications, and webhooks subscribed to comment.created receive the event).
comment_on_project
Create a new advanced checklist template for the current company. Sets the title and optional description; the template starts with no fields. Add sections and fields afterwards with update_conditional_checklist_template. Returns the created template.
create_conditional_checklist_template
Create an advanced checklist on a project by instantiating one of the company's checklist templates. Checklist items populate asynchronously — the response includes the new checklist (is_populating: true until the background job finishes) and the job id. Optionally assign company users (each assignee is notified and added to the project); ids outside the current company are ignored.
create_conditional_checklist_from_template
Create a kanban board (project pipeline) with named stages. On a label board each stage uses a project label (created if needed); on a stage board stages are named, colored columns that are not labels. Returns the board with its phases (tag fields only for label boards). Requires manager access. Use move_project_on_board to place projects afterward.
create_board
[beta] Create a new customer for the current company. Requires a `name`. Optionally attach `contacts` in the same call. Returns the created customer with its contacts. Fires customer.contact_created for each contact it creates. Beta — may change shape or behavior without the standard deprecation process.
create_customer
Create a document folder on a project, to group its Documents and Pages. Folders don't nest. Returns the created folder.
create_document_folder
Step 1 of uploading a local/binary document larger than the 1 MB inline limit of `create_project_document` (up to 30 MB). Returns a short-lived presigned S3 URL plus an `upload_id`. PUT the raw file bytes to `upload_url` using the returned `headers` (both Content-Type and Content-MD5 must be sent and must match), then call `create_project_document` with the `upload_id` (in place of `content` or `content_base64`) to attach the document. `checksum` is the base64-encoded MD5 of the exact bytes you will PUT.
create_document_upload
[beta] Create a draft invoice on a project. Omit proposal_id for a standalone invoice (contact fields prefill from the project). Pass proposal_id to raise an invoice from a proposal — optionally with invoice_type (deposit/progress/final) and allocation_in_cents. Requires CompanyCam invoicing. No funds are collected and no financial details are requested or returned. Status is not writable: this tool cannot send, open, void or settle an invoice. Add lines with create_invoice_line_item. Beta — may change shape or behavior without the standard deprecation process.
create_invoice
[beta] Add a line item to a draft or open invoice that is not being settled. Pass price_book_item_id to start from a catalogue item (request fields override catalogue defaults). Send unit_cost_in_cents only — unit_price_in_cents / customer_unit_price_in_cents are rejected. Optional position reorders after create. Requires CompanyCam invoicing. No funds are collected and no financial details are requested or returned. Beta — may change shape or behavior without the standard deprecation process.
create_invoice_line_item
Create a PAGE (a rich-text note or report) on a project. The title is rendered separately at the top of the page — do NOT repeat it as the first node of `tiptap_content`. `tiptap_content` is optional (omit it to create an empty page to fill in later) and must be a complete TipTap document object whose root is { "type": "doc", "content": [ ...top-level nodes... ] }. Wrap a photo grid, table, or notes block in a `section` node with the matching sectionType (the same block the editor's "+" menu inserts). Every section node must start with a heading holding its title, or the write is rejected. Otherwise author bare top-level nodes. Header/footer content does NOT belong in tiptap_content — it's a separate band at the top/bottom of the page that this tool cannot set. Any headerContent/footerContent included here is silently dropped. flexPhoto nodes need a real photoId from this project (look one up via list_project_photos); the server fills in the photo's image URLs from the id — never transcribe image URLs yourself. To edit the page afterwards, use update_page. This tool does not plan, draft, or improve content — it writes exactly what you provide. ---- ## TipTap schema for Page content Pages store content as TipTap JSON in `tiptap_content`. The root is always: { "type": "doc", "content": [ ...top-level nodes... ] } Generated from @companycam/tiptap-editor (see lib/tip_tap/schema.json). ### Document-level attrs Set these only through their dedicated tools where one exists (see each attr's note); otherwise leave them untouched. Never invent values. - `pageSettings` (object) — Root-level page rendering settings; not authored by agents - `photoCaptureSettings` (object) — Root-level photo capture defaults; not authored by agents - `headerContent` (object) — The page header band. A fixed three-column sub-document; set it with the write_page_header_footer tool, never by hand. Preserve it as-is on an edit. Any column may hold the company logo (see headerFooterColumn); band text supports bold/italic/underline/color only. May contain the literal page-number tokens {page} (current page) and {total} (total pages), which the editor substitutes per page at render time -- preserve them verbatim, never rewrite them into real numbers. - `footerContent` (object) — The page footer band. A fixed three-column sub-document; set it with the write_page_header_footer tool, never by hand. Preserve it as-is on an edit. Any column may hold the company logo (see headerFooterColumn); band text supports bold/italic/underline/color only. May contain the literal page-number tokens {page} (current page) and {total} (total pages), which the editor substitutes per page at render time -- preserve them verbatim, never rewrite them into real numbers. ### Node types you may emit - `doc` · content=(block | pageBreak)+ attrs: photoCaptureSettings: object, pageSettings: object, coverPageFooter: string - `paragraph` · group=block · content=inline* attrs: textAlign: one of ["left", "center", "right", "justify"] - `heading` · group=block · content=inline* attrs: textAlign: one of ["left", "center", "right", "justify"], level: integer - `text` · group=inline - `hardBreak` · group=inline - `pageBreak` · group=pageBreak note: A page break. Use when the user asks for a page break. Page breaks force the following content to start on a new page. Do NOT put pageBreaks inside of FlexGroup or a FlexPhoto. Page breaks go at the document root or directly inside a Section (after its heading), never nested deeper. A pageBreak directly inside a Section splits that Section across pages: content above it stays on the current page and content below it starts the next page, still inside the same Section. At the root they are most commonly used preceding a Section. - `bulletList` · group=block list · content=listItem+ - `orderedList` · group=block list · content=listItem+ attrs: start: integer, type: string - `listItem` · content=paragraph block* - `taskList` · group=block list · content=taskItem+ - `taskItem` · content=paragraph block* attrs: checked: boolean - `blockquote` · group=block · content=block+ - `horizontalRule` · group=block - `headerFooterColumn` · group=block · content=paragraph+ attrs: position: one of ["left", "center", "right"] note: One column of a page header/footer band. The band is a `doc` of exactly three headerFooterColumn nodes in order -- position "left", "center", then "right"; the editor stores each at its default alignment. Each column holds plain paragraphs (up to three lines) of text and, optionally, the company logo; band text supports bold, italic, underline, and color marks only -- no links, highlight, or nested layouts. The company logo is a bare image node (`{ "type": "image" }`) placed inside a paragraph, in any column of the header or footer, alone or beside text; the server supplies the logo itself. The logo fills the full height of the column, so a column holding it has just that one paragraph: text can go beside the logo, but not on other lines. The read tools show an existing logo the same way: keep that image node to keep the logo, and leave it out to remove it. Use the literal page-number tokens {page} (current page) and {total} (total pages) and preserve them verbatim. The band lives in the headerContent/footerContent doc attrs. - `image` · group=block note: The company logo in a page header or footer. It goes inside a paragraph of any headerFooterColumn, alone or beside text. Author it bare (`{ "type": "image" }`); the server supplies the logo itself. - `flexGroup` · group=block · content=flexPhoto* attrs: orientation: one of ["horizontal", "vertical"], size: one of ["xsmall", "small", "medium", "large"], aspectRatio: one of ["square", "original"], showPhotoText: boolean, showCaptureDetails: boolean, showTags: boolean, showPhotoLocation: boolean note: Container for one or more photo cards, each shown with optional caption text and metadata. flexPhoto children always stack vertically (one card per row). orientation positions the caption relative to the photo (horizontal: caption to the right of the photo; vertical: caption below). size controls photo dimensions. aspectRatio controls photo crop. flexGroup is the canonical container for any photo-bearing layout, including multi-column photo grids (configure via size + aspectRatio). When the user asks for photos paired with descriptions, captions, notes, or side-by-side text -- often phrased as a "table" of images with descriptions, or "images on one side and text on the other" -- reach for flexGroup with orientation: "horizontal" (caption right of photo) or "vertical" (caption below). Do NOT use a `table` node for image-plus-text layouts. - `flexPhoto` · group=block · content=flexPhoto_textSlot? attrs: photoId: string, description: string, capturedAt: string, capturedBy: string, tags: string, projectName: string note: A single photo card inside a flexGroup. Must reference a real photoId from the project; never invent ids. The server fills in the photo image automatically from photoId. - `flexPhoto_textSlot` · group=block · content=block+ photoCaptureDetailsV2? note: The visible text slot for a photo card. Include block content (paragraph, list, etc.) here when the user wants any visible text rendered alongside the photo, such as a caption, label, note, or callout. Omit entirely when photos should appear without text. - `photoCaptureDetailsV2` · group=block note: Auto-rendered capture metadata block (project name, captured by, captured at). Place inside a flexPhoto_textSlot when the user wants capture details visible alongside the photo caption. No attrs to set; visibility is controlled by the parent flexGroup.showCaptureDetails. - `table` · group=block · content=tableRow+ note: Tabular data layout. Use when the user asks for a table or structured rows-and-columns content. Children are tableRow. - `tableRow` · content=(tableCell | tableHeader)* note: A row inside a table. Children are tableHeader (for header rows) or tableCell (for body rows). - `tableHeader` · content=block+ attrs: colspan: integer, rowspan: integer, colwidth: string, align: string note: Header cell inside a tableRow. Use in the first row to label columns. Contents are plain block content (paragraph, list). Tables carry tabular text data only; place any photo-bearing layout (flexGroup, flexPhoto) at the top level of the document so it renders with room to expand. Sizing follows the same colwidth rule as tableCell -- set colwidth on every cell in the header row so widths sum to at most 624. Wrapping long labels: header text wider than the column expands the column and can push the table past the page margin. Split such a label across multiple paragraph children so it wraps within colwidth. Prefer word boundaries when the label has multiple words -- e.g. a 6-column table with "Problem or Service Required" uses two paragraphs, "Problem or" and "Service Required". When a single word is wider than its column, split it across paragraphs mid-word to keep the table within the page margin ("Mechanic" in a very narrow column may become "Mecha" and "nic"). Fitting the table inside the page takes priority over keeping a word intact. Short labels that already fit their column ("Order Date", "Start Time") stay in one paragraph. - `tableCell` · content=block+ attrs: colspan: integer, rowspan: integer, colwidth: string, align: string note: Body cell inside a tableRow. Contents are plain block content (paragraph, list, heading, blockquote). Tables carry tabular text data only; for photo grids or photo cards with captions, place a flexGroup at the top level of the document so it renders with room to expand. Sizing: colwidth is an array holding one integer (pixels), e.g. [156]. Set colwidth on every cell in a row so widths sum to at most 624 -- the Letter-format page content width between 1-inch margins. Default to even distribution: for a row with N cells, use colwidth: [floor(624/N)] on each. When the source has visibly uneven columns (e.g. a "Description" column wider than a "Qty" column), mirror those proportions and scale so widths still sum to at most 624. - `section` · group=block · content=(block | pageBreak)+ attrs: sectionType: one of ["photo", "notes", "table", "custom"], collapsed: boolean, origin: string note: Collapsible, draggable wrapper for one titled section of a page -- the same block a user inserts from the editor's "+" menu (Photo Section, Table Section, Notes Section). Content MUST start with a heading holding the section title: a Section starts with a heading (any level), and a section without a leading heading is invalid. H1 is reserved for the page title, so new Sections use H2 (heading level 2) for the title. The body after the heading is optional -- a heading-only Section is valid -- but add a paragraph or the typed block so the Section is not empty. Always give the heading real title text -- the editor auto-deletes a section whose heading and paragraphs are all empty. Set sectionType to match the content: "photo" (heading, optional intro paragraph, then a flexGroup -- ALWAYS include the flexGroup, left empty when no photos were requested or found; empty renders the editor's "Add photos" placeholder), "table" (heading, optional paragraph, then a table), "notes" (heading plus paragraphs/lists). Inside a photo section, copy the flexGroup attrs from the photo entry in the "Section presets" list unless the user asked for a specific layout. When the user asks for a section without saying what goes in it, mirror the editor presets exactly -- copy the matching entry from the "Section presets" list in this schema. Never author sectionType "custom" -- it appears on some existing documents and must be preserved as-is on edits. Never nest a section inside another section. nodeId is server-managed: omit it on new sections and preserve it byte-for-byte when editing an existing one. Wrap content in a section when the user asks for a section or the page already organizes its content in sections; otherwise author bare top-level nodes. - `coverPageV2` · group=block · content=coverPageV2_imageSectionSlot? coverPageV2_titleSectionSlot coverPageV2_logoSectionSlot note: The cover page -- the block the editor's Insert > Cover Page adds: cover photo, title, project address, date, and the company logo, name, address, and phone. Add one only when the user asks for a cover page, title page, or cover photo -- all mean this node. Before writing or rebuilding one, ALWAYS call read_cover_page_details (project_id for a page, template_id for a page template) in the same turn, and copy its returned values verbatim into the slots. This applies to every write that adds one, including creating a new page with write_page or create_page_from_template: the project_id you are creating the page in is the one to pass. Never write a value or placeholder token of your own -- header/footer tokens such as {Date} or {Company Name} do not fill in on a cover page. For a page template that does not exist yet: create it first with create_page_template (without the cover page), read it back with find_page_templates, call read_cover_page_details with its template_id, then add the cover page with update_page_template. It is always the first node of the document, and a document holds at most one. Content is ALWAYS all three slots in order: coverPageV2_imageSectionSlot, coverPageV2_titleSectionSlot, coverPageV2_logoSectionSlot -- never leave one out. Leave its attrs out. On an edit, keep an existing cover page first and preserve its slots and photos as they are; change its text only when the user asks. To remove the cover page, drop the node. - `coverPageV2_imageSectionSlot` · group=block · content=photo note: The cover photo slot. Always include it, on a page or a page template. Holds exactly one photo node: `{ "type": "photo", "attrs": { "ratio": "original", "showControls": false, "id": "<photo id>" } }`. Use cover_photo_id from read_cover_page_details, or another real project photo id the user asked for. Leave id out when there is none and on a page template -- keep the photo node itself, since an empty one is where the cover photo goes. Never write a src -- the server fills the image from the id. - `coverPageV2_titleSectionSlot` · group=block · content=heading paragraph paragraph note: The title slot. Holds exactly three nodes in order: a level-1 heading with the title, then a paragraph with project_address, then a paragraph with date, each value from read_cover_page_details; mark both paragraphs' text bold. On a page the heading is the page title; on a page template it is the title value read_cover_page_details returns. - `coverPageV2_logoSectionSlot` · group=block · content=photo paragraph paragraph paragraph note: The company slot. Holds exactly four nodes in order: the company logo as a photo node `{ "type": "photo", "attrs": { "ratio": "original", "showControls": false, "isCoverPageLogo": true } }`, then three paragraphs with company_name (bold), company_address, and company_phone from read_cover_page_details. Never write the logo's src -- the server fills in the company logo. ### Inline marks `bold`, `italic`, `link` (attrs: href), `underline`, `highlight` (attrs: color) ### Section presets When the user asks for a section without saying what goes in it, mirror the editor presets exactly -- emit a `section` with the matching `sectionType` whose content is the preset children below. Section titles are H2 (heading level 2) and H1 is reserved for the page title, so set the leading heading's level to 2 even when a preset below shows another level: - photo: [{"type":"heading","attrs":{"level":2}},{"type":"paragraph"},{"type":"flexGroup","attrs":{"aspectRatio":"square","showPhotoText":true,"showCaptureDetails":true,"showTags":true,"showPhotoLocation":true,"orientation":"horizontal","size":"medium"}}] - notes: [{"type":"heading","attrs":{"level":2}},{"type":"paragraph"}] - table: [{"type":"heading","attrs":{"level":2}},{"type":"paragraph"},{"type":"table","content":[{"type":"tableRow","content":[{"type":"tableHeader","content":[{"type":"paragraph"}]},{"type":"tableHeader","content":[{"type":"paragraph"}]},{"type":"tableHeader","content":[{"type":"paragraph"}]}]},{"type":"tableRow","content":[{"type":"tableCell","content":[{"type":"paragraph"}]},{"type":"tableCell","content":[{"type":"paragraph"}]},{"type":"tableCell","content":[{"type":"paragraph"}]}]},{"type":"tableRow","content":[{"type":"tableCell","content":[{"type":"paragraph"}]},{"type":"tableCell","content":[{"type":"paragraph"}]},{"type":"tableCell","content":[{"type":"paragraph"}]}]}]}] ### Example body A top-level heading of any level starts a Section in the editor. Use H2 for Section titles; H1 is reserved for the page title. { "type": "doc", "content": [ { "type": "heading", "attrs": { "level": 2 }, "content": [ { "type": "text", "text": "Today's progress" } ] }, { "type": "paragraph", "content": [ { "type": "text", "text": "Framing finished on the east wall. " }, { "type": "text", "text": "Inspector signed off at 2pm.", "marks": [ { "type": "italic" } ] } ] }, { "type": "orderedList", "content": [ { "type": "listItem", "content": [ { "type": "paragraph", "content": [ { "type": "text", "text": "Drywall delivery tomorrow." } ] } ] }, { "type": "listItem", "content": [ { "type": "paragraph", "content": [ { "type": "text", "text": "Crew of 4 starting at 7am." } ] } ] } ] } ] } ### Example section A `section` node wrapping a titled block of content: { "type": "doc", "content": [ { "type": "section", "attrs": { "sectionType": "notes" }, "content": [ { "type": "heading", "attrs": { "level": 2 }, "content": [ { "type": "text", "text": "Site conditions" } ] }, { "type": "paragraph", "content": [ { "type": "text", "text": "Dry and clear; roof deck accessible." } ] } ] } ] }
create_page
Step 1 of uploading a local/binary photo — use this instead of a public image_url. Returns a short-lived presigned S3 URL plus an `upload_id`. PUT the raw image bytes to `upload_url` using the returned `headers` (both Content-Type and Content-MD5 must be sent and must match), then call `create_project_photo` with the `upload_id` (in place of `image_url`) to register the photo. `checksum` is the base64-encoded MD5 of the exact bytes you will PUT.
create_photo_upload
[beta] Create an active item in the company price book. Requires price_book_items:write. Stays off until an admin or manager turns on "Let admins and managers update the Price Book through AI tools" under Integrations → CompanyCam MCP. Changing your company price book needs manager access — standard users cannot update the catalogue through AI tools. Set unit_cost_in_cents and markup; customer_unit_price_in_cents is derived and cannot be written. If you need to add more than one item, use create_price_book_items instead of calling this once per item. Beta — may change shape or behavior without the standard deprecation process.
create_price_book_item
[beta] Create many active items in the company price book in one call. Use this instead of calling create_price_book_item once per item. Accepts up to 100 items. An item that fails — a missing name, a duplicate name, or a rejected price field — fails on its own and does not roll back the rest of the batch. Requires price_book_items:write. Stays off until an admin or manager turns on "Let admins and managers update the Price Book through AI tools" under Integrations → CompanyCam MCP. Changing your company price book needs manager access — standard users cannot update the catalogue through AI tools. Set unit_cost_in_cents and markup on each item; customer_unit_price_in_cents is derived and cannot be written. Beta — may change shape or behavior without the standard deprecation process.
create_price_book_items
Find-or-update: returns the existing same-company project with the same name and jobsite address (updating it, and restoring it if deleted) instead of creating a duplicate; otherwise creates a new project. The address fields are the jobsite/property being worked on, never a person's home or current location. Requires at least one of `name` or `street_address_1`. Optional `customer_id` also requires the `customers:write` scope. Returns the project. Fires project.created (and project.contact_created when primary_contact is given). If the company has a connected storage integration (Dropbox, Google Drive, SharePoint), creating a project also creates a folder there named with the project name and jobsite address.
create_project
Create an empty, one-off advanced checklist on a project, with no template. The checklist is ready immediately (is_populating: false) and has no fields yet. Add sections and fields with update_project_checklist_structure. To start from one of the company's templates instead, use create_conditional_checklist_from_template. Optionally assign company users when creating: each assignee is notified and added to the project; ids outside the current company are ignored. Only available when one-off checklist authoring is enabled for the account; otherwise this returns an error and creates nothing. Not idempotent: calling it twice creates two checklists, so check list_project_checklists before retrying. Returns the checklist in the same envelope as get_project_checklist.
create_project_checklist
Attach a document (PDF, text, etc.) to a project. Provide `content` as a UTF-8 string (it will be base64-encoded for upload) or `content_base64` for binary files, up to 1 MB. For a larger file (up to 30 MB), call `create_document_upload` first, PUT the bytes to its presigned URL, then pass the returned `upload_id` here instead. `filename` should include the extension. Processing is asynchronous; it fires the document.created webhook and may sync the document to the project's connected third-party integrations. Set `access_limited` to true to mark the document Restricted (visible only to its creator and users who can view all project documents). For a rich-text note or report authored in CompanyCam, use `create_page`.
create_project_document
Create a project group on the current company — this bundles projects, not users (to group users instead, use create_user_group). Name must be unique within the company. Returns the created group; add projects to it with add_project_to_group.
create_project_group
Create a collaborator invitation on a project. When redeemed, it offers ANOTHER company access to the project (view, add content, comment, use to-dos). Returns the new invitation's status and metadata. The `invite_url` is intentionally omitted: it embeds a capability token that grants project access, so it is never returned to an agent. Deliver the invitation through CompanyCam's own UI/email rather than handing the link to a model. Once issued, the invitation cannot be recalled.
create_project_invitation
Upload a photo to a project. Provide EITHER `image_url` (a publicly fetchable http(s) URL the photo is downloaded from asynchronously) OR `upload_id` from create_photo_upload (after PUTting local bytes to its presigned URL) — use `upload_id` to upload a local file with no public host. Optional: captured_at timestamp, description, jobsite coordinates (where on the property the photo was taken; defaults to the image's EXIF location), and tags (created on the company if they don't already exist). Returns the created photo with all variant URLs.
create_project_photo
Create a single assignable task in a project's Tasks section and optionally assign it to one or more company users, in one call. This is the modern per-project "Tasks" surface (one task per call). It is NOT the checklist/punch-list tool — use create_conditional_checklist_from_template to add a multi-item checklist to a project from one of your company's templates. `details` is the task text (1-5000 chars). `assignees` are the IDs of users in your own company; each must be assignable to the project or the call fails. `position` (optional) sets the 0-indexed slot in the list. Fires the project_task.created webhook; assignees are also added to the project and are emailed and push-notified.
create_project_task
[beta] Create an unaccepted proposal on a project. Contact and address fields prefill from the project when omitted. Returns the created proposal. Status is not writable — do not send, accept, decline, sign, or void. Unit price is derived; set costs on line items with create_proposal_line_item. A proposal carries one discount, taken off the subtotal before tax: either discount_percentage or discount_amount_in_cents, never both non-zero. A fixed amount is refused on create, before line items exist; set it with update_proposal. Beta — may change shape or behavior without the standard deprecation process.
create_proposal
[beta] Add a line item to an unaccepted proposal. Pass price_book_item_id to start from a catalogue item (request fields override catalogue defaults). Send unit_cost_in_cents only — unit_price_in_cents / customer_unit_price_in_cents are rejected. Optional position reorders after create. Refused once the proposal is accepted or signed. Beta — may change shape or behavior without the standard deprecation process.
create_proposal_line_item
Create a company-scoped tag (media taxonomy). Idempotent: if a tag with the same display value already exists on this company, the existing one is returned. Returns the tag.
create_tag
Create a webhook subscription on the current company and return the new webhook. The HMAC signing token is not returned here (`token` is null) and a delivery `Authorization` header cannot be set here, and there is no web app setting for it. To set a delivery `Authorization` header, use `PATCH /public_api/v1/webhooks/{id}` (REST API) with the new webhook's id (`{"webhook":{"authorization_header":"..."}}`); to rotate the signing token, `PATCH` with `{"webhook":{"token":"..."}}`. Do not use the legacy `/v2` API. `get_webhook`/`list_webhooks` report only whether a header is set. `subscription_url` must be HTTPS in production and must not resolve to an internal or private address (loopback, link-local, private ranges); invalid URLs are rejected. `scopes` is a list of event names from the public scope vocabulary (e.g. "project.created", "photo.tag_added", "comment.created"). Wildcards like "project.*" and "*" are supported.
create_webhook
Fetch a kanban board by ID, including stages (phases) and the projects currently in each stage. A project appears in only one stage. Label boards return tag_id/tag_name/tag_color per stage; stage boards return name, color and is_default_entry instead (stages are not labels). Use projects_per_phase to cap how many projects are returned per stage (default 50, max 100); list_board_phase_projects pages the rest.
get_board
Fetch one CLASSIC (legacy) checklist template by ID, including its sections, tasks, and sub-tasks (questions) with answer_type and answer_options. Requires the templates plan. Use list_legacy_checklist_templates to find the id. This is not an advanced checklist template; use list_conditional_checklist_templates for those. The purpose of this tool is to read a classic template so it can be rebuilt as an advanced template with create_conditional_checklist_template. answer_options lists the possible answers for a sub-task; on checklist instances, answer_choices are zero-based indexes into answer_options.
get_legacy_checklist_template
Fetch the current authenticated session's company: id, name, status, address and logo URLs, plus documents_enabled (true when Documents is available to the current user and Reports are fully hidden for them).
get_current_company
Return the authenticated user (the OAuth token's resource owner). Use this first to confirm which user (and via `company_id`, which company) the session is operating against.
get_current_user
[beta] Fetch a single customer by ID, including address and notes. Beta — may change shape or behavior without the standard deprecation process.
get_customer
Fetch one DOCUMENT (an uploaded PDF/file attachment) by ID: name, content type, size, creator, project and folder IDs, `access_limited` (Restricted), and a download URL (the latest signed version's URL when the document has been signed). For a rich-text note or report authored in CompanyCam, use get_page.
get_document
Read the extracted TEXT of one DOCUMENT (an uploaded PDF/file attachment) by ID, in character slices. Use this when the question is what a document says; get_document returns only its metadata and download URL. The text is machine-extracted, not the original layout, so tables may flatten and formatting is lost. Returns `content`, `offset`, `limit`, `total_length`, `truncated`, and `extraction_status`. When `truncated` is true, call again with offset = offset + limit to read the next slice. `extraction_status` is `extracted` (text available), `pending_or_empty` (text not extracted yet, or the file has no text such as a scanned image), or `unsupported` (file type is not text-extractable). For a rich-text note or report authored in CompanyCam, use get_page.
get_document_content
Fetch one document folder by ID. Folders group Documents and Pages within a project; check each record's `folder_id` from list_project_documents / list_project_pages to see what's inside, since neither tool takes a folder filter.
get_document_folder
[beta] Fetch one invoice by ID, including line items, totals, and contact details. Prefer this over listing line items separately when you need the full document. Requires CompanyCam invoicing. No funds are collected and no financial details are requested or returned. Status is not writable: this tool cannot send, open, void or settle an invoice. Beta — may change shape or behavior without the standard deprecation process.
get_invoice
[beta] Fetch a single line item on an invoice by ID. Requires CompanyCam invoicing. No funds are collected and no financial details are requested or returned. Beta — may change shape or behavior without the standard deprecation process.
get_invoice_line_item
Fetch one PAGE (a rich-text note or report authored in CompanyCam) by ID. Pass `include` to opt into heavier fields: `raw_content` (plain-text projection) or `tiptap_content` (raw, unstable TipTap document). For an uploaded PDF/file, use get_document.
get_page
Fetch a single photo by ID with all URI variants and metadata. Pass include='annotations' to also get the text and sticker keys a user drew on the photo. Prefer that over view_photo when the question is what the photo *says* rather than what it looks like — reading the annotation text costs a fraction of downloading the image. If you need more than one photo, use `get_photos` instead to fetch them in a single round trip.
get_photo
Fetch metadata for multiple photos by ID in one call. Use this instead of calling get_photo once per photo — e.g. after list_project_photos or list_photos — to look up several photos in a single round trip. Accepts up to 20 photo IDs per call (duplicates are only fetched once); `include` applies to every photo in the batch. Returns a per-photo result list — a photo that doesn't exist or isn't visible to you fails individually without blocking the rest of the batch.
get_photos
[beta] Fetch one active price book (catalogue) item by ID: name, description, unit cost, markup, derived customer unit price, unit type, and tax flag. Available to all roles. Beta — may change shape or behavior without the standard deprecation process.
get_price_book_item
Fetch one PROJECT (a job/site) by ID: name, status, archived/public flags, jobsite address, jobsite coordinates and geofence (the property being worked on, not a person's location), primary contact, cover photo, description, photo and document counts, and share URLs. Not a document or page (see get_document, get_page).
get_project
Fetch one advanced checklist on a project by ID, including its field tree, per-field completion, live progress counts, and any photos or videos attached to completed fields. Use list_checklists to find what changed company-wide, or list_project_checklists to discover a checklist_id on a known project. Instance fields include a `media` array. Each item is `type: "photo"` or `type: "video"` plus the same fetchable URLs as get_photo / get_video (`uris` on photos, `playback_url` on videos). That is how you get inspection photos off a photo-required field — they are not on list_project_checklists. Prefer get_photo with include=annotations when you need the text drawn on a photo rather than the image itself. Large checklists cannot be returned in full — a 100-question checklist is roughly 33,000 tokens, and field media adds more. By default this returns the most detailed view that fits a size budget and reports in `meta.notice` when it had to narrow, so a read never fails for size alone. Check `meta.view` before concluding a question is absent: `answers` omits unanswered questions, `outstanding` omits answered and hidden-branch questions, `applicable` omits hidden branches, and `summary` omits the fields entirely. If you need the photos and auto returned `summary`, request view=answers or view=outstanding explicitly. For "what's still left on this job", request view=outstanding. That returns only questions that currently apply given the checklist's show/hide rules and have not been answered — not the unanswered fields on branches this job never reached. list_project_checklists reports applicable_fields_count / outstanding_fields_count for the same distinction as counts. To set answers, use update_project_checklist with a batch of field ids. Read completed_by_name and completed_at first and do not overwrite a CompanyCam user's completion without confirming. To change the instance title or description, use update_project_checklist_details. To add, edit, move, or delete fields on the instance, use update_project_checklist_structure.
get_project_checklist
Fetch one CLASSIC (legacy) checklist on a project by ID, including its sections, tasks, sub-tasks (questions), and recorded answers and completion state. Use list_project_legacy_checklists or list_legacy_checklists to find the id. This is not an advanced checklist; use get_project_checklist for those. The purpose of this tool is to read a classic checklist so it can be rebuilt as an advanced checklist with create_conditional_checklist_template or create_conditional_checklist_from_template. Each sub-task has an answer_type and answer_options (the possible answers); answer_choices are zero-based indexes into answer_options, not the option text, and answer_text holds free-text answers.
get_project_legacy_checklist
Fetch the customer linked to a project, including address, notes, and contacts. A project with no linked customer returns `data: null` — that is a normal answer, not an error, so there is nothing to retry.
get_project_customer
Fetch a single project group by ID, including name, description, status, and project count. A project group bundles projects; for a group of users use get_user_group.
get_project_group
CompanyCam ChatGPT Plugin FAQ
How the directory, categories and Discoverability Score work.
Read the methodologyHow do I improve CompanyCam's ChatGPT Plugin discoverability?
The levers are the listing surface agents actually read: names, descriptions, keywords, tool metadata, and registry health. Which lever matters depends on where discovery breaks, which is what continuous measurement shows.
What are CompanyCam alternatives on ChatGPT?
As of 2026-11-20, CompanyCam competes with 3montovano CRM, A4B CMMS, AI Dispatcher by FieldCamp, Apex Rental Pro, BlueSuite, Bobibo, BoltWright, Cabinet Project and 49 more in ChatGPT Field Service Management Software, ranked by public Discoverability Score.
Where is this profile measured?
This profile uses the geography attached to the latest public registry snapshot: US. Locale tags are intentionally omitted.