What is diagram?
diagram is a Claude Code mod by korenka that adds a Diagram button that draws the flow or architecture from Claude's latest response in its own pane tab, saved as SVG and draw.io. It is listed under Interface and has no public popularity data yet.
| Type | Claude code mod |
|---|---|
| Author | korenka |
| Repository | korenka/claude-harness |
| Popularity | no public popularity data yet |
| Category | Interface |
| Security | High risk (56/100) |
| Rank | #3,730 of 5,272 Claude Code mods; #1,551 in Interface |
| License | MIT |
| Last commit | 2026-10-13 |
| Hooks | 5 events |
| Sees | only the events it hooks |
| Draws on | AbovePrompt, Pane |
Is diagram safe to install?
diagram scores 56/100 on the MCPMods security check: High risk. Things to review: starts processes outside the sandbox; writes files or config; reads environment variables. It passed 3 of 8 checks.
Mod footprints come from claude plugin validate, which lists every event a mod hooks and every API call it makes without running it.
| Check | Result |
|---|---|
| No network access | Pass |
| Starts processes outside the sandbox | Review |
| Writes files or config | Review |
| Reads environment variables | Review |
| Only sees the events it hooks | Pass |
| Drives Claude (model calls, prompts) | Note |
| Validator reported problems | Note |
| Has an open-source license | Pass |
API calls in its code
| Call | What it means |
|---|---|
$.clock.now | |
$.command.register | |
$.env.get | reads env vars |
$.fs.write | writes files |
$.model.complete | drives Claude |
$.model.fork | drives Claude |
$.process.run | runs processes |
$.session.cwd | |
$.session.messages | reads the transcript |
$.ui.copy | draws interface |
$.ui.invalidate | draws interface |
$.ui.log | draws interface |
$.ui.open | draws interface |
$.ui.resolve | draws interface |
How do I install diagram?
- Check your Claude Code version. Mods need Claude Code 2.1.287 or later and are on by default.
claude --version
- Read its footprint. This lists every event it hooks and every API call it makes, without running it.
git clone https://github.com/korenka/claude-harness claude plugin validate ./claude-harness/mods/diagram
- Install it. Run this in your terminal, then /reload-plugins in an open session.
git clone https://github.com/korenka/claude-harness && claude --plugin-dir ./claude-harness/mods/diagram
Or ask Claude to set it up
Paste this into Claude Code and it installs diagram for you:
Install the "diagram" Claude Code mod from https://github.com/korenka/claude-harness/tree/main/mods/diagram. Run claude plugin validate on it first and show me what it can access before you enable it.
More from korenka/claude-harness
This repository has 6 Claude Code mods on MCPMods.
- agent-dock: Agents side panel with run, pause, stop and open-in-tab controls, plus a context window bar with a Compact button (Low risk)
- blind-spots: Adds a Blind Spots button above the prompt that asks Claude to list the engineering blind spots in the current plan or d (Low risk)
- help-offers: Turns the 'How I can help' list at the end of an answer into buttons that start a subagent for each offer (Low risk)
- rephrase: Adds a Rephrase button that rewrites the draft in the prompt box with a small model, so the LLM gets a clearer, structur (High risk)
- tldr: Adds a TLDR button that summarizes Claude's latest response in a side pane (Low risk)