ThoughtDAG: Editable context graph for organizing LLM conversations

What it does

ThoughtDAG is an open-source AI conversation canvas that lets users edit the context sent to language models. Instead of linear chat threads, it organizes conversations as a directed acyclic graph where users can branch discussions, explore side questions, and selectively bring past conversations into new prompts. Users can remove unwanted context routes, inspect what tokens will be sent to the model, and reuse conclusions from earlier threads without losing the original discussions.

Who it is for

The tool targets users who conduct complex, nonlinear research or problem-solving with LLMs. It suits researchers comparing multiple hypotheses, analysts exploring alternative explanations, or anyone whose thinking branches into tangents that later need to be integrated or excluded from final requests. The desktop app indexes local conversation records that users connect to it.

Pricing

The site does not show prices. ThoughtDAG is open-source under the MIT license.

How it stands out

Most AI chat interfaces assume linear conversation flow. ThoughtDAG treats thinking as a graph where branches are first-class objects. Users can explore a tangent question, see the results on the canvas alongside the main thread, then decide which parts feed into the next prompt. The tool visibly shows users what context—measured in tokens—will actually reach the model, addressing "context pollution" where unrelated chat history pollutes answers. It lets users edit context without deleting conversation history, preserving reasoning chains for reference while controlling what enters each request.

What a founder should check

First, verify switching costs from existing workflows. ThoughtDAG requires users to intentionally connect local chat records; it does not automatically restore chat history or access conversations stored only in web accounts. Examine whether the friction of manual connection and graph editing outweighs the benefit of fine-grained context control for typical users.

Second, test the moat around editable context graphs. Commercial AI tools may add graph-based context management to their own chat interfaces, removing the advantage of a standalone app. Check whether being open-source strengthens or weakens defensibility as larger platforms adopt similar features.

Third, explore pricing pressure. The tool is free and open-source. A commercial competitor could charge for hosted graph management or premium token-preview features. Assess what revenue model—if any—the creators envision and whether free distribution limits growth incentives.

Thinking of building something like this?

Every launch here is a competitor to somebody's idea. If yours is close, check it against the market before you build: the Full Check names the rivals, the prices and the gaps.