flow

Parse the session trace file and generate a Mermaid flowchart of skills and agents.

Updated Feb 8, 2026
One-click install
npx skills add https://github.com/ckallum/calsuite --skill flow-ckallum
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: flow
Source: https://github.com/ckallum/calsuite/tree/main/.claude/skills/flow
Command: npx skills add https://github.com/ckallum/calsuite --skill flow-ckallum

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Generate a clear, human-readable diagram of the current session's workflow by converting trace data into a Mermaid flowchart that shows the order of skills and agents, including parallel branches and collapsed repeats.

Core Features & Use Cases

  • Mermaid flowchart generation: produces a readable diagram of session activity.
  • Ordering and concurrency: captures sequential execution and parallel tasks across skills and agents.
  • Repeat-collapsing: hides repeated dispatches to keep diagrams concise.
  • Use Case: debugging and optimizing session workflows by visualizing task order and parallelism.

Quick Start

Run the flow skill to generate a live Mermaid diagram of the current session's workflow.

Frequently Asked Questions about flow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I visualize a development session workflow as a Mermaid flowchart?

You can visualize a development session workflow by parsing the session trace file and mapping the sequence of dispatched skills and agents into a Mermaid flowchart. This approach converts trace data into a readable diagram showing task order and parallelism.

Can I generate a Mermaid diagram that shows parallel tasks and repeated dispatches?

Yes, generating a Mermaid diagram from session trace data captures both sequential execution and parallel branches across skills and agents. It also includes repeat-collapsing logic to hide redundant dispatches and keep the visual output concise.

What is a JSONL trace file and when do I need it for workflow visualization?

A JSONL trace file records the sequential execution log of skills and agents within a session. You need it for workflow visualization when you want to debug or optimize task order, concurrency, and repeated dispatches by mapping them into a structured diagram.

Do I need a specific trace file path to map session concurrency and task order?

Yes, workflow visualization requires access to the trace file located at .claude/flow-trace-${CLAUDE_SESSION_ID}.jsonl. The skill parses this specific file to identify and map the sequence, concurrency, and repetition of dispatched tasks.

What is the best way to debug parallel task execution across multiple agents?

The best way to debug parallel task execution is converting session trace data into a Mermaid flowchart. This visualizes the exact ordering and concurrency of dispatched skills and agents, making it easier to identify inefficiencies in the workflow.

Are there limitations when visualizing repeated skill dispatches in a Mermaid diagram?

A key limitation when visualizing repeated dispatches is diagram clutter, which this approach mitigates by collapsing repeated task sequences. However, it relies entirely on the availability and integrity of the JSONL trace file to produce accurate flowcharts.