Knowledge, skills, notes, and tasks

Each organization has an isolated workspace that gives people and agents shared context without making chat history or an external document system the source of truth.

Ask Stella about a saved record

Open a saved Knowledge document, a Personal or Workspace skill, or an existing note or task, then open Stella. Notes and Tasks provide an Ask Stella control in the editor; returning from Stella preserves unsaved edits. The selected record identifies the source for the next question. Skill selection includes its exact saved version. Recent skill runs and System templates do not bind a saved skill. Unsaved editor text is not included in this selection.

Canvass checks current source access on each turn. Closing the record or removing it from the current view clears the selection. The shared context reader must be enabled for the organization and source kind; selection alone does not enable it or grant access. These sources use the same existing protected context reads exposed through the API, CLI and MCP gateway. Complete related-history and semantic search coverage are still being expanded.

Author with AI

Select Add knowledge or Create skill to open a persisted, organization-scoped assistant thread. While the tenant-bound chat and its first message are being committed, every launch control is disabled and reports Creating governed chat…. A second click cannot create a duplicate thread. The first message records the complete authoring intent and safety boundary rather than a generic placeholder. The assistant:

1. Reads the matching governed authoring guide.
2. Reads the tenant workspace to find overlaps and conflicts.
3. Asks focused questions about intent, sources, triggers, expected output, permissions, and failure behavior.
4. Drafts readable Markdown and, for a skill, proposes realistic test prompts.
5. Runs skill prompts in an isolated, read-only model simulation and displays individual findings.
6. Shows the exact proposed document in a preview card.
7. Waits for a natural person to select Save draft.

Saving runs through Temporal and Fabric Platform's governed mutation pipeline. The assistant cannot activate a record, expand its own permissions, publish content, enable delivery, or send outreach. Assistant-authored records always enter the workspace as drafts.

On the first real model turn, the assistant proposes a short history title through the same governed chat-persistence adapter. When the operator later supplies the actual topic or workflow, the assistant may replace that launch title with a materially more specific one. Every title is checked against the current organization, user, and chat binding before PostgreSQL accepts it. After each turn, the chat header and history are refetched so they show the latest AI-derived context and actual response. If title generation does not complete, the full initial request remains as the deterministic fallback; Canvass never fabricates a completed response or progress event. A turn that can update history allows one additional bounded tool step; other turns retain the standard eight-step limit.

The chat composer accepts one Markdown, MDX, text, or JSON source file up to 50 KB. Canvass hashes the text in the browser, labels it as untrusted user-supplied input, and keeps it attached if sending fails so the operator can retry. The source is not treated as verified knowledge merely because it was uploaded. When sent, its bounded text and hash are persisted with the organization-scoped chat message in PostgreSQL. It does not enter the Knowledge library until the assistant proposes an exact draft, a natural person approves Save draft, and the record is later activated through the governed workspace lifecycle.

Author manually

Select Write manually to open the structured editor immediately. Knowledge and skill editors provide:

• Preview and Write modes;
• rendered headings, lists, tables, links, emphasis, block quotes, and code;
• workspace or personal scope plus hierarchical folders;
• source metadata, content hashes, AI provider/model/prompt provenance, tags, and skill policy fields;
• structured skill test cases;
• a document viewer with Viewer, Editor, Copy link, and Download actions;
• a contextual assistant bar that can receive bounded Markdown, text, or JSON source files.

Manual creates, updates, and deletes use the same governed mutation path as assistant proposals.

Navigate the resource workspace

Knowledge and Skills use a file-oriented, three-pane workspace:

• Use the compact + action beside the library title to start guided authoring. The larger empty-state actions remain available when a library has no records.
• Workspace and Personal roots keep organization context separate from the current operator's private drafts.
• Hierarchical folders use the record's governed folder value, and documents appear as Markdown files.
• Selecting a folder shows its description, direct files, and a rendered index.md document when the folder contains one.
• The Skills workspace also exposes read-only System packages for Add knowledge and Create skill. Their skill.md files describe the exact authoring workflow and safety boundaries.
• The assistant composer remains bound to the selected file, folder, or system guide. Its context chip makes that binding visible before a message or source file is submitted.

Search filters titles, descriptions, and folder paths while preserving the matching folder hierarchy. Folder and document URLs are shareable within the authenticated organization.

Storage and retrieval

Knowledge and skills are tenant-scoped operational records backed by PostgreSQL. The Cloudflare application reads current snapshots; saves use the existing Temporal and Fabric Platform Host path. gtm.upsert_knowledge, gtm.upsert_skill and their lifecycle operations retain durable events and current source revisions. Storage adapters and derived indexes never replace that authority.

Open a saved Knowledge document and use Ask Stella in the page toolbar to include that record in the next assistant turn. Selecting another document changes the next turn's context. Folders and authoring guides are not saved Knowledge records, and unsaved editor text is not automatically included as the saved source.

The workspace reader setting controls how Stella receives that saved content. With shared context enabled for Knowledge in Settings → Access → Workspace context, Stella reads the current permitted record through the shared reader, with source references and bounded continuation. Otherwise, the legacy path includes up to 8,000 body characters, 2,000 summary characters and ten source URLs as untrusted source data in the model prompt. This applies both to the default legacy reader and to shared context with Knowledge left unchecked: neither setting turns off selected-document context. Pausing workspace context blocks selected Knowledge reads instead of using that fallback. Both paths enforce personal-record ownership; archived or unavailable records cannot supply saved content.

The library search filters titles, descriptions and folder paths. The shared reader provides exact and keyword reads; its general semantic search and background consumers remain disabled. A separate legacy Knowledge vector-search tool exists behind the deployment flag GTM_CHAT_VECTOR_SEARCH=1; that legacy flag does not enable shared semantic search. Approved company-document retrieval is also separate and does not establish semantic coverage for every Knowledge record, chat or customer interaction. External agents still need their approved read grants.

Files supplied to an authoring chat are bounded source inputs, not knowledge records. They become canonical knowledge only after the assistant proposes an exact draft and a natural person selects Save draft. Binary asset storage and retrieval indexes remain adapters; neither is allowed to become an alternative source of business truth.

Current authoring boundaries

Canvass intentionally does not give chat a general shell, filesystem, raw SQL, or unrestricted web browser. The current authoring tools can read tenant workspace and strategy projections, read the system authoring guides, run bounded skill simulations, and propose approval-gated knowledge, skill, ICP, and artifact drafts. Uploaded text sources are size-bounded, hashed in the browser, explicitly delimited as untrusted data, and retain their supplied-source metadata. A URL pasted into chat is not claimed as fetched or verified; governed URL fetching, reusable binary assets, and conflict-aware OpenKnowledge editing remain separate future adapters.

Next: Agents, Artifacts, and Models and gateways.