Catch the fleeting thought
Capture first without breaking concentration. The inbox keeps the original text, voice, image, link, time, and context until the person decides what deserves a future.
AI-first private alpha
A second brain that thinks with you.
Treefiny is an AI-first thinking system and living slip-box for learning, research, recall, and creation. Capture any question, source, fact, or fleeting idea. Specialist AI helps you clarify it, connect it, challenge it, remember it, and turn it into original work—while you remain the author of the graph.
Private-alpha notes only. No unrelated campaigns.
- User-owned graph
- Source-aware AI
- Reversible by design
ROOT QUESTIONWhat do I want to understand?Yours · now
SOURCEAttention shapes memory.From source · p. 42
BRIDGE THOUGHTRecall can be a creative act.AI inference · review
SYNTHESISA path worth writing.3 claims · 5 sources
The product thesis
Knowledge should do more than wait to be found.
Most knowledge tools stop after storage and search. Treefiny asks what evidence supports a claim, reconnects an old idea to a new project, prepares concepts for recall, reveals contradictions, and helps a developed branch become original work. The person remains the author; the system supplies structure, memory, and momentum.
Inspired by Maggie Appleton’s Programmable Notes—and extended with the slip-box tradition into a universal, graph-based product for thinking, learning, and creation.
The slip-box, rebuilt for AI
Your ideas should converse—not disappear into folders.
Treefiny combines the slip-box principles of stable addresses, focused permanent thoughts, deliberate links, and living entry points with multimodal capture, source-aware AI, semantic connection discovery, spaced recall, and programmable review.
Read with the source attached
Keep quotations, paraphrases, claims, page or timestamp locations, bibliographic data, and retrieval dates visibly connected instead of blending source and interpretation.
Make one idea durable
Turn understanding into a stable permanent thought in the person’s own language. AI can coach atomicity and prepare a split, but the person approves the idea and its boundaries.
Connect it with a reason
Link thoughts through support, contradiction, continuation, example, dependency, question, or derivation. Every AI suggestion explains the relationship it sees.
Build living entry points
Use sequences and structure notes to guide a course, topic, argument, or project. One thought can appear in many maps without being copied into disconnected folders.
Write from the conversation
Let AI assemble approved claims, evidence, tensions, and sources into alternative outlines and drafts while every sentence can still resolve to the graph underneath.
Grounded in the Niklas Luhmann Archive, Zettelkasten Method principles, Sönke Ahrens’ How to Take Smart Notes, and the Writing Scientist’s research-oriented introduction. Treefiny adapts the method; it does not copy paper constraints or source prose.
A calm operating surface
One graph.
Every useful view.
A focused branch for today. A source trail when rigor matters. A recall queue when memory matters. A synthesis canvas when the thinking is ready to become work. Nothing drifts into a disconnected copy.
ROOT QUESTION
When does remembering become creating?
12 thoughts · 5 sources · 2 unknownsCLAIMRecall reconstructs, not replays.Supported by 2 sources
EVIDENCERetrieval changes the memory retrieved.From source · accepted
TENSIONBut repetition can reduce variation.Unresolved · review
CONNECTION SCOUT
This source may qualify your claim because it separates fluent recall from reconstructive recall.
Accept linkNot relevant
One system, many minds
For anyone trying to understand something.
A student and a scientist need different workflows, not different foundations. Treefiny gives every person the same durable graph, then adapts the prompts, views, memory rhythm, and outputs to the work in front of them.
Learn anything deeply
Turn lessons, books, lectures, and questions into a concept graph. Use teach-back prompts, practice questions, and spaced recall to move from exposure to durable understanding.
Research with evidence
Build a literature trail from a real question. Keep sources, claims, evidence, counterevidence, quotations, and uncertainty connected so every synthesis can show where it came from.
Remember what matters
Choose what deserves to stay available. Treefiny schedules active recall, resurfaces relevant context, and adapts review to what is becoming strong, weak, or newly useful.
Map complex ideas
Move fluidly between a focused branch, a mind map, an outline, and a wider knowledge graph. Every view reflects the same underlying ideas and relationships.
Create from knowledge
Develop connected material into a study guide, research brief, essay, plan, explanation, or original synthesis without losing sources, tensions, or open questions.
Think through decisions
Compare options against explicit criteria, constraints, risks, and unknowns. Preserve the reasoning behind a decision and define what evidence should trigger a review.
The product surface
One continuous path from capture to creation.
Treefiny is not eight disconnected tools in one subscription. Each surface reads and writes the same user-owned knowledge model, so a source can become evidence, a concept can become a memory, and a mature branch can become useful work without losing context.
Universal capture
Start with text, a question, a link, a document, an image, audio, or a fleeting thought. AI transcribes and prepares the inbox while preserving the raw capture and its origin.
AI-native slip-box
Turn source notes and the user’s own permanent thoughts into stable, addressable ideas connected by explained relationships, sequences, and structure maps.
Research and fact layer
Separate what a source says from what the user believes and what the system infers. Track citations, support, contradiction, confidence, and the date a fact was checked.
Memory engine
Generate recall prompts from real knowledge, schedule spaced repetition, reveal why an item returned, and reconnect each memory to the context that makes it useful.
Visual thinking
Branch, cluster, sequence, and compare ideas on calm canvases. Mind maps are navigable views of the graph—not isolated drawings that drift away from the knowledge.
Workflow studio
Create plain-language programs with triggers, conditions, actions, and review. Begin with trusted templates, then adapt and share workflows without needing to code.
AI thinking partner
Ask precise questions, find connections, challenge claims, compare perspectives, and draft source-aware synthesis. Every suggestion explains why it appeared.
Output studio
Turn a developed graph into useful work: explanations, flashcards, briefs, outlines, essays, plans, decisions, and next actions that remain linked to their foundations.
The knowledge model
A graph that knows the difference between an idea and a fact.
Everything is a durable, typed object with provenance, time, status, ownership, and relationships. Hierarchy keeps a thread navigable; cross-links let knowledge behave like a graph. No important reasoning is flattened into an anonymous block of text.
- Inbox item
- A raw, temporary capture waiting to be discarded, deferred, linked, or developed.
- Source note
- What a source says at a precise page, section, timestamp, or location.
- Permanent thought
- One durable, reusable idea explained in the person’s own language.
- Structure note
- A curated entry point that orders existing thoughts without duplicating them.
- Question
- What the person wants to learn, resolve, or decide.
- Concept
- A durable idea, definition, model, or principle.
- Claim
- A statement that can be supported, challenged, or revised.
- Fact
- A checkable assertion with source, date, and confidence.
- Source
- A book, paper, page, person, recording, or dataset.
- Evidence
- Material that supports, contradicts, or qualifies a claim.
- Memory
- Something intentionally prepared for future recall.
- Connection
- A typed relationship with an explanation of why it matters.
- Synthesis
- A reviewable understanding drawn from connected material.
- Decision
- A conclusion with rationale, alternatives, and review conditions.
- Action
- A concrete next move derived from the thinking.
- Unknown
- A visible gap, tension, assumption, or unresolved dependency.
Inspectable graph intelligence
Do not just retrieve an answer. Reveal the path.
Treefiny combines reliable structural extraction with bounded semantic inference. The graph can answer questions, trace relationships, reveal useful neighborhoods, and show what changed—while keeping source facts, user authorship, and AI proposals distinct.
Structure before inference
Treefiny reads reliable structure—citations, headings, links, authors, dates, and user-authored relationships—before asking AI to interpret meaning. Models propose what cannot be extracted directly.
Every edge tells the truth
Connections distinguish what you authored, what a source states, what a rule derived, and what AI inferred. Certainty and review state stay separate, visible, and correctable.
Ask, trace, explain
Ask a question against a focused subgraph, trace how two ideas connect, or explain why a thought matters. Answers return inspectable objects and paths—not only fluent prose.
Find neighborhoods and bridges
Discover thematic regions, central ideas, surprising cross-field bridges, orphaned thoughts, and ambiguous links. Graph signals guide attention without pretending connectivity equals truth.
See what changed
New and updated sources produce a reviewable graph diff. Treefiny shows what was added, contradicted, relinked, or superseded without silently rewriting permanent thoughts.
Retrieve with a visible path
Lexical search, semantic similarity, metadata, time, and graph traversal work together. Every item in an AI context packet explains why it was selected and how it reaches the active question.
This architecture learns from the open-source Graphify project, particularly its structural-first graph construction, explicit relationship provenance, community analysis, incremental updates, and query, path, and explain operations. Treefiny adapts those principles to human learning and thinking rather than codebase mapping.
Programmable thinking
Workflows people can read, own, and change.
Every program has a trigger, conditions, actions, and a human review step. Begin with a proven learning or research practice, inspect exactly what it does, and adapt it in ordinary language instead of surrendering control to invisible automation.
- PRESERVE the raw input and its origin
- PROPOSE discard, defer, connect, or develop
- REVIEW before anything becomes permanent
- ASK for the idea in the person’s own words
- CHECK that it is focused, sourced, and understandable later
- SUGGEST where it joins the existing conversation
- ASK for an explanation in the learner’s own words
- CREATE recall prompts only after review
- RESURFACE them on an adaptive schedule
- REQUEST a source and supporting evidence
- FIND possible counterevidence and related context
- REVIEW every proposed relationship
- SHOW the relevant ideas side by side
- EXPLAIN the shared concept or relationship
- LET the person link, dismiss, or defer
- DRAFT the current understanding and its sources
- SURFACE tensions, unknowns, and alternatives
- PROPOSE a conclusion or next action for approval
The experience loop
Thinking that compounds instead of disappearing.
The loop is simple enough for a first thought and powerful enough for years of work. Each pass strengthens the graph and makes future recall, research, and creation more relevant—without rewarding accumulation for its own sake.
- 01
Capture
Save the thought or source before it disappears.
- 02
Connect
Place it near what gives it meaning.
- 03
Develop
Question, test, compare, and expand it.
- 04
Recall
Bring it back when time or context makes it useful.
- 05
Create
Turn understanding into an explanation, decision, or action.
AI with bounded authority
AI is the interface. The graph is the memory. You are the author.
Treefiny can ask, retrieve, compare, challenge, and draft. It must never manufacture a source, silently rewrite the graph, or hide uncertainty behind a confident answer. Useful intelligence begins with legible boundaries.
Every change is reviewable.
Accept, edit, dismiss, defer, disable, or undo AI-proposed connections and content.
Every answer shows its footing.
User statements, quoted sources, system inference, and speculation remain visibly distinct.
Knowledge stays within its boundary.
Private graphs are scoped, sharing is opt-in, and export, deletion, and model-use policies are explicit.
Reflection wins over engagement.
The product optimizes for clarity and durable understanding—not streaks, noise, or endless capture.
The adjacent landscape
Learn from strong products. Keep the thesis distinct.
Treefiny sits beside connected-note systems, visual research spaces, AI notebooks, and learning tools. Each proves an important behavior; Treefiny joins those lessons in one longitudinal reasoning graph built around provenance, agency, recall, and creation.
Capacities
Connected, typed objects and flexible views.
Treefiny extends the object model into explicit reasoning relationships, source boundaries, recall, and claim-level synthesis lineage.
Explore ↗Heptabase
Visual learning and research through cards, sources, and whiteboards.
Treefiny keeps every visual arrangement synchronized with a typed graph, programmable reflection, memory, and reviewable AI patches.
Explore ↗Tana
Structured nodes, supertags, AI, commands, and events.
Treefiny puts evidence, uncertainty, relationship provenance, calm reflection, and reversible authorship at the center of automation.
Explore ↗NotebookLM
Source-grounded AI research, citations, and multimodal explanations.
Treefiny develops durable thought objects and cross-project paths that connect sources to a person’s evolving ideas, decisions, and creations.
Explore ↗RemNote
Notes, source annotation, AI study tools, flashcards, and spaced repetition.
Treefiny connects recall to questions, claims, evidence, projects, conceptual transfer, and the wider reasoning graph.
Explore ↗Obsidian
Private local Markdown, links, graph and canvas views, and extensibility.
Treefiny pairs durable ownership with an opinionated AI-first thinking method, typed provenance, and explainable graph operations.
Explore ↗Comparisons reflect each product’s public materials as reviewed on August 26, 2026. They describe emphasis rather than exhaustive capability or private roadmaps.
The build standard
Fast enough for a fleeting thought. Rigorous enough for a serious conclusion.
A great Treefiny experience makes capture effortless, relationships understandable, provenance inspectable, recall purposeful, and AI actions reversible. It works with keyboard, touch, and assistive technology; remains calm from phone to wide screen; and turns the same knowledge graph into the right view for the moment.
Private alpha
Bring your thinking with you.
Join the Treefiny waitlist for product previews, research invitations, and early-access openings. This list is only for Treefiny.