The reason-commons skill now saves working reasoning with Claude Code's tracked file-editing tools so rewind can restore the tree with the conversation. After restoration, rerunning the existing views helper refreshes the diagrams and records reversals of accepted reasoning as new revisions while retaining earlier history.
Pull request approval checks now allow explanatory paragraphs alongside approval references, avoiding formatting-only failures while retaining approval and scope validation.
Pull requests can adjust supported diagram display settings while the reasoning remains checked against the tree. Merging into main requires a successful reasoning check and an up-to-date branch, including for administrators.
Reason Commons now presents each requested project change, including “fix it,” as a typed proposal with explicit tree updates, waits for typed acceptance, and records the identity, version, scope, recording time and exact reply before implementation. Pull requests show checked reasoning together in one diagram per affected tree, explicitly identify unresolved connections, keep the complete change list collapsed and carry recorded approval; completion reports retain the reasoning trace, next-step recommendations have their own format, and desktop diagrams identify each node's tree and type.
When you ask the reason-commons skill to show a tree, it now draws the real diagram in the chat, with drag to pan and buttons, pinch or ⌘/Ctrl+scroll to zoom, instead of retelling the tree as a list of chains. The diagram, its node counts and the "Not shown" line are generated straight from the saved reasoning, so smaller models such as Haiku get them right too.
Every pull request to this repository now says which part of our reasoning tree it serves and which tree updates it makes, and an automated check confirms that list against the tree, so work can't quietly drift away from the goal. The reason-commons skill works the same way for any project that keeps its tree in its own repository, and a new blog post walks through how the change came about.
A new glossary explains every Logical Thinking Process term and acronym you meet in the guide, the books and the skills' output — CRT, UDE, injection, evaporating cloud and the rest — each under its own heading so it can be linked to directly, with a small sketch of each tree worked through one running example, e.g. reasoncommons.com/glossary#current-reality-tree.
In the Claude apps, the reason-commons skill now draws each turn's change-status diagram as a rendered Mermaid flowchart inside the conversation, with proposed links dashed and accepted, proposed and disputed nodes outlined in different colours. Terminals and other plain-text chats keep the ASCII diagrams.
Reasoning sessions now put an ASCII status card above the direction choices to summarize the latest proposal, acceptance, dispute, or lack of change, and the final view choice lists available trees with node counts and a direct prompt for viewing one.
skills/README.md now leads with a simple installation path and copy-and-paste prompts for both conversational reasoning and source-to-project conversion. It explains that reason-commons is a candidate successor to ltp-project, while ltp-project remains the route to an app-importable file and ltp-visualize still needs a clone of this repository. ltp-project no longer asks for an app checkout to validate its output; it labels the file unvalidated instead, and a standalone check is tracked as a follow-up.
Flowershow
