Codepact

Treescribe Getting Started

Returns an orientation guide to Treescribe: what the service is, how to get an account and sign in, and what the tools unlock once authenticated. Callable without authentication. Use this when: you have just discovered this server and are not signed in yet, or want a from-zero overview before calling anything else. Inputs: none. Response fields: markdown. Example: {} → { "markdown": "# Welcome to Treescribe\n…" }

Treescribe Guides Read

Returns Treescribe's long-form usage guide for one tool space (or the concepts primer). Use this when: about to use a space's tools for the first time in a session — e.g. read document-chat before your first chat-send, organisation before writing knowledge records. Inputs: guide — one of concepts, organisation, templates, documents, document-chat, document-ops. Response fields: guide, markdown. Example: { "guide": "document-chat" } → { "guide": "document-chat", "markdown": "# Document Chat\n

Treescribe Org Whoami

Returns the authenticated Treescribe user's identity, organisation, practice domains (jurisdictions), document count, the web app base URL, and the organisation's monthly AI allowance. Use this when: orienting at the start of a session, labeling the connection, or checking which org's data other tools will act on. Call it before anything else. Inputs: none. Response fields: userId, name, username, email, orgId, domains[] each { id, title } (the account's practice domains — document listing, t

Treescribe Org Knowledge Add

Creates one organisation knowledge record — an atomic firm fact that informs document drafting. Use this when: storing a fact the org's documents should reflect (standard positions, firm details, negotiated defaults). One fact per record. ALWAYS call treescribe-org-knowledge-list with a query first to avoid near-duplicates; prefer treescribe-org-knowledge-update over re-adding. Inputs: heading (required — the searchable claim), knowledge (the substance), description, relevanceOrg (when the fac

Treescribe Org Knowledge Update

Updates fields of an existing organisation knowledge record. Use this when: a stored fact changed or needs refinement — prefer this over adding a near-duplicate record. Get the _id from treescribe-org-knowledge-list. Inputs: _id (required); any of heading, description, relevanceOrg, knowledge, date — only the fields you pass change. Response fields: _id. Example: { "_id": "k1", "knowledge": "Liability capped at 6 months of fees" } → { "_id": "k1" }

Treescribe Org Knowledge List

Lists (or fetches one of) the organisation's knowledge records, with substring filtering and pagination. Use this when: checking what the org already knows before adding a record (dedup), finding a record's _id for update/delete, or surveying stored facts. Matching is plain case-insensitive substring over the record fields — no relevance ranking. Inputs: _id (fetch exactly one record); query (substring filter); limit (default 20) and offset for paging. Response fields: with _id → item; otherw

Treescribe Org Knowledge Delete

Permanently deletes one organisation knowledge record. This cannot be undone. Use this when: a stored fact is wrong or obsolete and should stop informing drafting. If the fact merely changed, prefer treescribe-org-knowledge-update. Inputs: _id (required — from treescribe-org-knowledge-list). Only records of your own organisation can be deleted. Response fields: _id, deleted: true. Example: { "_id": "k1" } → { "_id": "k1", "deleted": true }

Treescribe Templates List

Lists the legal document templates available to create documents from, limited to the account's practice domain(s) (jurisdiction) when set. Use this when: choosing a template before treescribe-template-document-create, or answering what kinds of documents Treescribe can produce. Pass search to narrow by name/heading. Inputs: search (optional substring, case-insensitive over name/heading/guide). Response fields: templates[] each { templateId, name, heading, guide, domains[] (domain ids), workf

Treescribe Template Document Create

Creates a new legal document from a template, in the account's practice domain (jurisdiction) — e.g. an Australia account always gets an Australian document. The document starts with the template's defaults; progressing it (answering its decisions) happens through the document chat. Use this when: the user wants a new document. First call treescribe-templates-list to get valid template names. Inputs: template (name from templates-list, required), headingUser (optional document title), subHeadi

Treescribe Document List

Finds and lists the user's legal documents, newest first, scoped to the account's practice domain(s) (jurisdiction) when set. Use this when: locating an existing document before reading or chatting about it. Not for creating documents — that is treescribe-template-document-create. Inputs: query (optional substring over document titles), folderId (optional scope to one folder), limit (default 20). Documents in archived folders are excluded unless folderId targets one explicitly. Response field

Treescribe Document Read

Returns a document's finished legal text — what its sections render to right now — as markdown. Use this when: summarizing, reviewing, comparing, or quoting a document's actual content. Not for progressing the document — use treescribe-document-chat-send for that. Inputs: documentId (from treescribe-document-list or treescribe-template-document-create); version (optional saved snapshot version — omit for the live current state). Response fields: documentId, version, markdown (the full legal t

Treescribe Document Steps Get

Exports a document's full step questionnaire — every decision, its current answer, section settings, properties and parties — in the exact format treescribe-document-steps-apply accepts. Use this when: inspecting what is answerable deterministically, checking current answers and valid decision names, or building a bulk payload for steps-apply. For interactive completion prefer treescribe-document-chat-send — chat is the default way to progress a document. Inputs: documentId (required; from tre

Treescribe Document Steps Apply

Applies a structured steps payload to a document in one bulk operation: properties, topic/definition answers (by decision name), party names, title/subtitle and sources. Only the fields you include change. Use this when: you already know exactly what to change — e.g. filling answers from a data source or replaying a saved questionnaire. Chat (treescribe-document-chat-send) is the default way to progress a document; steps are the deterministic path. Call treescribe-document-steps-get first for c

Treescribe Document Chat Read

Reads a document's chat: recent messages, the current thing needing attention, and a status telling you whether to wait or act. Creates the document's chat thread on first call (which starts its guided interview). Use this when: starting a document's interview (first call creates the chat thread), after treescribe-template-document-create, checking state or recovering after a disconnect, and polling during long runs. treescribe-document-chat-send returns the updated state directly, so you norma

Treescribe Document Chat Send

Sends a message into a document's chat and waits for the agent's response: free text, or a reply to the pending form from treescribe-document-chat-read. Use this when: answering the current form (pass reply items referencing its artefactIds) or giving the document agent an instruction in free text. Never send while chat-read reports processing. Inputs: documentId; content (free text) and/or reply — array of { messageId (the form's messageId), artefactId, artefactGroupId?, value } where value i