Deep institutional knowledge for AI workflows.

Build a living repository for everything you read — PDFs, web pages, spreadsheets, video — each one prepared for search the moment it lands. Pull the exact passage you need with a single query, and add your own notes and writing as you go.

A search in the workspace: a result fragment clicked open in its source on the right, the matched passage lit while the surrounding text dims

One search · the exact passage, lit in its source

Trusted by researchers at

UC BerkeleyNYUUC Davis

Seamless collection

Save websites, files, and video transcripts into one library, straight from your browser.

Precise recollection

Every upload is prepared for semantic search: enter the concept you need and pull up its exact location within your documents.

Structured AI

Mark up your data with rich annotations, citations, task-specific threads, and more. See real results from AI by binding it to your own perspective.

Fully collaborative

Shared collections with everyone's insights discoverable and kept in context.

Surface the specifics of past work with ease

Create living repositories for everything that goes into complex, long-running research projects. Search for what you remember and pull up anything in its precise original context.

Attention, Curation, and the Shape of Thought

Memory is not a passive archive. What we retain is shaped, at the moment of encoding, by the questions we bring to it.

Every act of saving is also an act of selection, and what you choose to keep is itself an act of interpretation — a first draft of the meaning you will later find there.

We tend to imagine recall as retrieval, as if the past sat unchanged in a drawer. But the file you reopen has been quietly edited by everything you filed beside it.

The act of curation — deciding what to keep, and how to arrange it — is not separate from thinking but constitutive of it: the structure we impose becomes the substrate on which new connections form.

Creative insight, on this account, is less a spark than a recombination: the novel connection was latent in material already deliberately set side by side.

Over months and years this compounds. The archive you build becomes a testament to the complexity of your thought, reflecting the judgement that is uniquely yours.

To curate well, then, is to think ahead about one’s future self — the reader who will return, months later, needing exactly this.

Search in Cognition

Attention, Curation, and the Shape of Thought3 sections

what you choose to keep is itself an act of interpretation.

the structure we impose becomes the substrate on which new connections form.

The archive you build becomes a testament to the complexity of your thought.

Structure your material according to your perspective

Annotate passages, leave notes in the margin, and jot down longer ideas with citations to the exact location of their supporing evidence. What accumulates is a map of your own connections and judgement instead of a flat document dump.

Put AI to use for your own thinking

Start with your own material and the specifics of how you use it, then layer in AI as it makes sense. Follow connections you drew months apart, surface conclusions you made in passing, or pull up and synthesize everything you have on a given topic, in seconds. Connect your own AI tools through our MCP server. Team plans have optional built-in chat.

You · May 29, 2026

Curation isn’t storage — choosing and arranging IS the interpretive act. The shelf is a first draft.

Linked sections

You · May 27, 2026

Cf. Dewey: reflection as the deliberate re-ordering of experience — the order is the argument.

in source Attention, Curation, and the Shape of Thought

Memory as Reconstruction

Added by Jake · yesterday

Remembering has long been modelled as playback: a record retrieved intact from storage. The experimental record reads otherwise.

Recall is performance, not playback — each act of recall is a small act of revision — the remembered scene is rebuilt around the concerns of the present.

In this sense what is stored is not the experience itself but the most recent reconstruction of it. The archive is alive, and every visit leaves fingerprints.

For shared work the implication is immediate: a collaborator returning to your notes reconstructs them through their own questions — division of attention becomes division of memory.

None of this argues against keeping records. It argues for keeping them deliberately — knowing the keeping itself will do quiet editorial work.

Linked sections· drawn Apr 14

Memory as Reconstruction

each act of recall is a small act of revision — the remembered scene is rebuilt around the concerns of the present

Attention, Curation, and the Shape of Thought

We tend to imagine recall as retrieval, as if the past sat unchanged in a drawer. But the file you reopen has been quietly edited by everything you filed beside it.

does recalling a memory change the memory itself

Attention, Curation, and the Shape of Thought
1 matching section
- We tend to imagine recall as retrieval, as if the past sat unchanged in a drawer. But the file you reopen has been quietly edited by everything you filed beside it.

~/cognition

~/cognitionmcp · thought-process ✓

Collaborative perspective in one place

A team thread in the Thought Process workspace — the feed of posts, attached source passages, and the collaborators on it

Thought Process makes it easy to harness the insights that otherwise get buried in shared drives, email attachments, and browser bookmarks. Keep everyone's contributions in one place with task-specific threads, where everything is prepared for nuanced recollection and AI integration.

See Thought Process for teams

Your work first, AI second

Connect coding agents, external RAG pipelines, or custom agents through our MCP server. The model is yours, meaning its data handling follows your own plan with that provider.

Your library is yours

Export your sources, notes, and writing anytime. No lock-in, nothing held hostage.

Privacy is primary

Your material is yours alone and never used to train models. Shared workspaces are visible only to people you invite.

Frequently asked questions

No. Your uploaded sources, notes, and writing are yours alone and are never used to train models. Integration with third party AI is up to you and governed by your usage agreement with those models.

Whichever you bring. Thought Process exposes your library through an MCP server, so you can connect coding agents like Claude Code, Codex, Cursor, a RAG pipeline, or any agent loop — and swap models whenever you like.

Yes. Your library exports in full — sources, annotations, links, and writing — so there is no lock-in and nothing is held hostage.

Only the people you invite. Collaborators see the collections you share with them; everything else stays private to you.

PDFs, Word docs, spreadsheets, web pages, video transcripts, and your own notes. You can upload through the web app or install the browser extension to save with one click as you browse. Every upload is chunked and embedded, making it ready for semantic search and AI integration.