How to use Claude to track habits in JSON
The author turns conversational reflections into structured JSON, views them in a local character-sheet dashboard, and exports edited context for later sessions.
news.ycombinator.com
Prompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Use [my reflections] and [existing character-sheet JSON schema] to identify active quests, skills, recurring obstacles, and recent achievements. Update [existing JSON context] in that schema for my local dashboard. Return the JSON so I can inspect, edit, and export it for our next session.
Reach people exploring AI agents
See placements →What happened
The author reports that the conversational feedback and gamified dashboard made journaling and habit tracking fun and helped previously overlooked progress feel substantial.
Reported by samholmes2 on Hacker NewsHow the agent did it
- Brain dump the things to track into Claude.
- Have the model infer quests, skills and recurring patterns.
- Ask the model to update the structured JSON data.
Where a person stepped in
The author pastes the output into the dashboard and can edit entries inline before exporting the context.
A person approved each step before the agent continued.
The character-sheet creator wanted more control over context carried between conversations than built-in project memory provided. Their approach asks an LLM to update a structured JSON file that can be visualised in a local dashboard. They describe using Claude or other models, including local models for personal material.
The workflow begins with a brain dump of whatever the author wants to track. The model infers recurring patterns from that conversation, such as active quests, skills, recurring enemies, and recent achievements. The author then asks it to “update my data.” The resulting structured JSON is pasted into the character-sheet interface, where a local HTML dashboard renders the information without accounts or a server.
The author can edit entries inline, export the updated context, and bring it to the next session. They say watching XP bars and graphs change makes progress they might otherwise forget feel substantial, and that the conversational feedback made journaling and habit tracking fun for them. They built the free, open-source interface iteratively with Claude Code Pro over a couple of months during evenings and weekends. Who it's for: people who like gamified reflection and want editable, portable context between AI conversations.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · samholmes2
Frequently asked questions
How does context move between sessions?
The model outputs structured JSON, which the author pastes into the character-sheet UI, edits if needed, exports, and brings into the next session.
Does the dashboard require an account or server?
The author says the dashboard renders locally as an HTML file, with no accounts or server needed.
More details
Remaining steps
- Paste the JSON into the local character-sheet dashboard.
- Edit entries inline and export context for the next session.
About this account
Claude infers recurring patterns from conversation and outputs structured JSON when asked.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=47987227. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: samholmes2 on Hacker News
Source published 2 May 2026
Source checked 14 Oct 2026
Tools and topics
Claude · JSON · character-sheet · HTML
Habit tracking · JSON · Journaling
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.