Launching today

claude-devtools

Launching today

See everything Claude Code hides from your terminal

65 followers

See everything Claude Code hides from your terminal

65 followers

Visit website

LLM Developer Tools

Not another Claude Code GUI wrapper. claude-devtools doesn't run or modify Claude Code — it reads the raw session logs already on your machine and reconstructs everything the CLI hides. Every file path Read, every tool called, every diff Applied, every token consumed — structured into a visual timeline with per-turn context attribution, compaction visualization, subagent execution trees, and custom notification triggers. Works with every session you've ever run. Open source, runs locally.

Free

Launch tags:Open Source•Developer Tools•Artificial Intelligence

Launch Team / Built With

Anima Playground — AI with an Eye for Design

AI with an Eye for Design

Promoted

Maker

📌

Hey PH! 👋 I built claude-devtools because I was frustrated with a specific problem: Claude Code stopped showing me what it's doing. Recent updates replaced detailed tool output with opaque summaries — "Read 3 files", "Edited 2 files" — no paths, no content, no diffs. The context usage became a vague progress bar. The only alternative is `--verbose`, which dumps thousands of lines of raw JSON. There's no middle ground. I shared this on Reddit as a weekend project to scratch my own itch, and the response was insane—over 100k views, 500+ upvotes, and 1,000+ downloads in just 48 hours. Clearly, we all hate coding blind! 😂 I tried every Claude Code GUI out there (Conductor, Craft Agents, Vibe Kanban, etc.). None of them solved this because they all *wrap* Claude Code. They inject prompts and only show sessions run through their own UI. If you ran a quick command in your terminal, it doesn't exist to them. claude-devtools takes the exact opposite approach: ⚡ It doesn't touch Claude Code at all. ⚡ It just reads the raw session logs already sitting at `~/.claude/` and reconstructs the full execution trace. Every file read (with syntax highlighting), every edit (with inline diffs), every tool call paired with its result, and every subagent's full execution tree. 🔥 The feature I'm most proud of is Visible Context Reconstruction: It reverse-engineers what's actually eating tokens in your context window, broken down across 7 categories per turn, showing exactly how context fills and compresses over a session. It's 100% free, open source, zero network calls, and works on macOS/Linux/Windows/Docker. What's your biggest frustration with the Claude CLI right now? I'd love to hear your thoughts, and I'll be here all day answering questions!

Report

8h ago

Hey @matt_1398 congrats on the launch!
> What's your biggest frustration with the Claude CLI right now?

I also like to keep a "tight leash" on Claude/Codex, but besides seeing what has changed I also want to see why it did that. The session log should reflect that, maybe you can find a way to highlight when Claude changes course and why it did that!

Report

5h ago

Maker

@janschutte

Thanks Jan! 🚀 That’s a great question, and it's something I think about a lot too.

Right now, claude-devtools surfaces those extended thinking blocks alongside the tool traces, so you can actually read the reasoning behind each step instead of just seeing the result. When you pair that with the full tool call chain, you can usually piece together why it took a specific path.

But you’re hitting on something deeper: explicitly flagging the "pivot" moments. Whether it’s abandoning an approach or shifting strategy after a tool error, having a dedicated "decision point" marker in the timeline would be a game-changer for observability.

I’m definitely adding that to the roadmap—really appreciate the killer suggestion!

Report

2h ago

Hey @matt_1398

This is interesting, thanks for building this.

It would be helpful if there could be a convenient progress bar view of the session and weekly % consumption to avoid using /usage

Report

1h ago