Workspace knowledge
The wiki your drafts and agents read first
Stateless assistants forget your positioning every session. This GTM knowledge graph holds your profile, ICP, offers, and voice. Its pinned voice, brand, and positioning pages ground drafts and chats with Oxygents, OXYGEN's in-app AI agents, by default.
AI without your context starts from nothing. You paste the same positioning into a prompt, get something almost right, tweak it, and paste it again next week. What is missing is durable company knowledge that lives in the workspace and gets read without anyone attaching it.
Two shapes share one store. A structured core: the six-section company profile plus the voice, brand, and positioning pages you pin as official. And linked pages for personas, proof points, research, meeting notes, and campaign learnings. Beneath both sit immutable sources and a log naming who, or which run, made each change.
Example searches
Ask for it the way you'd say it
Each example is a request in plain words, with what it returns underneath.
Write it down once
record our positioning, our three offers, and who we disqualify, then use it everywhere
A pinned positioning page grounds your drafts; offers and disqualifiers reach a column you map them into.
What did the campaign teach us
summarize what the last campaign's replies and results taught us
Synthesis turns the campaign's stats and reply classifications into a learnings page that states what it read. Auto-filing reply excerpts as sources is off for now.
Who changed this and when
show me every revision of the voice page, who made each change, and when
Each revision keeps its text, author, and time, and the log names who or which run made it, so a wording change is reviewable.
Plays
Three motions you can run this week
Each one is a chain of OXYGEN primitives: the same hosted objects your workspace already has, composed.
Seed the wiki, then fill the stubs
A new workspace starts with a real page skeleton instead of an empty text box nobody opens.
- 1Seed the starter set: draft ICP and Offers pages.
- 2Fill the ICP and Offers stubs, then write a voice page and approve its pin: it grounds your drafts.
oxygen knowledge seedClose the learning loop
What a campaign's replies taught you becomes a learnings page the next brief is written from.
- 1Run synthesis on the campaign: it reads the sequence's stats and how its replies were classified.
- 2Save the preview you reviewed as the campaign-learnings page the next brief starts from.
oxygen knowledge synthesizeGate the copy customers read
Working notes update freely; official brand and positioning edits wait for an approval.
- 1Let ordinary pages write through as logged, revertible revisions.
- 2Review official voice and positioning edits in the proposal queue before they land.
oxygen knowledge proposals
Capabilities
What you get
Grounded by default
The official voice, brand, and positioning pins go into your message, sequence, and post drafts and every Oxygent chat, without anyone attaching them. Drafts don't search your other pages yet: that waits until a human signs off on how well it works.
Provenance on the output
Drafts and AI cells record which page revisions grounded them. When a draft says something odd, you can open the exact version of the page behind it.
A risk-tiered write gate
Working pages auto-write as revisions anyone can revert. Official voice, brand, positioning, and pinned playbooks change only when a proposal is approved, and each decision is recorded.
A local mirror
Sync the wiki down to files, edit it in the editor you already use, and push it back. The workspace stays the source of truth and your terminal stays the interface.
Boundaries
What Knowledge Graph does not own
The knowledge graph describes kinds of customers. Actual customers are Records. Working datasets are Tables, message bodies are Messages, and an installable executable motion is a Recipe rather than a page you write.
Data sources
Data sources and CRM connections
ExaReads your site's pages for the first company research, which cites each one as a source on its research page.
FirecrawlReads the pages you point at, and your homepage when the first company research cannot read it with Exa.
LinkupOn your own Linkup key, a second research path when one engine's view of the web is too thin to write from.
- PaParallel
On your own Parallel key, web search steered by a research objective, for a page that needs more sources.
Use managed credits where available, or connect a supported provider key on an eligible paid plan. Your provider bills its calls; OXYGEN adds a per-call platform fee. Signed-in CRM connections such as HubSpot have no per-call fee. Connected sending accounts reserve monthly credits. See pricing and supported integrations.
Limits
Where this stops
- Grounding is only as good as what you wrote. An empty wiki produces generic copy very politely, and the first hour spent on ICP, offers, and voice is the highest-leverage hour in the workspace.
- Synthesis is one low-cost model call over evidence you already have, not a research project. It distills your campaign stats and reply classifications; it does not find new market facts for you.
- Automatic grounding is narrow today. Drafts get only the official pins; Oxygent chats also get a summary of your company profile and pages that match your question. Retrieval for drafts and auto-filing of replies wait on a human review.
FAQ
Questions people ask first
Why not keep all this in a docs tool?
What gets injected automatically?
Does an agent need permission to write here?
Can I edit this from my own editor?
What counts as a source?
Does writing a page cost credits?
Try Knowledge Graph in your own workspace
Sign up, get a working workspace with a one-time credit grant, and run the first play from the web app, the CLI, or your AI assistant over MCP.