Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

make-trace (Flowtrace trace builder)

Turns any source that describes how a task gets done — a SKILL.md, chat log, runbook, or plain prose — into a runnable Flowtrace DAG and drives the full run lifecycle to verify it.

AutomationAdvanced★ 481⑂ 31AI score 8/10Last updated: Jun 9, 2026

What it does

Lifts a source into a trace: a folder holding a DAG that both a human and an AI read while work runs.

  • Scaffolds with flowtrace init <slug>
  • Extracts the steps hidden in the source, fills trace.json#steps and from_steps to build a real fan-in/fan-out DAG instead of a flattened line
  • Enforces an independent faithfulness pass: every step present, every edge real, nothing dropped or invented
  • Writes a per-step contract at steps/<id>/STEP.md with reads/writes frontmatter
  • Places inputs as plain files in resources/, then drives run new → step running → done --asset → reply → deliverable done
  • Covers reuse (new run, swapped inputs, same plan) and steering (flowtrace show --downstream <id> to re-run only stale dependents)
  • Includes a "precipitate a completed run" mode: generalize a finished task by subtraction into a reusable method

Who it's for

  • Developers already using or evaluating Flowtrace
  • Agent-orchestration engineers who want repeatable methods as visible graphs
  • Teams turning tribal runbooks into steerable, inspectable pipelines

Examples

  1. Hand it an existing SKILL.md and get a 6–8 node trace, verified with flowtrace show --fmt mermaid
  2. Build a resume-optimization trace once, then swap files in resources/ to rerun it for a new candidate
  3. Take a task you just finished and precipitate it into a generalized trace, stripping product names and one-off values
  4. Change your mind about a mid-graph step and re-run only its transitive dependents, then re-close the deliverable

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/AIScientists-Dev/Flowtrace/HEAD/skills/make-trace/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/make-trace folder from the GitHub repo AIScientists-Dev/Flowtrace into my ~/.claude/skills/make-trace/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/AIScientists-Dev/Flowtrace.git && mkdir -p ~/.claude/skills && cp -r Flowtrace/skills/make-trace ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal and clone the repo: git clone https://github.com/AIScientists-Dev/Flowtrace.git
  2. Create the skills folder: mkdir -p ~/.claude/skills
  3. Copy the skill: cp -r Flowtrace/skills/make-trace ~/.claude/skills/ — make sure the bundled references/CLI.md comes along
  4. Install the flowtrace CLI: you need Node and Rust; run ./scripts/install.sh from the repo root to build the web UI plus CLI and symlink the binary into ~/.local/bin (first build takes a few minutes)
  5. Verify with flowtrace --help; add ~/.local/bin to your PATH if the command isn't found
  6. Restart Claude Code, then ask something like "make a trace from this document" to trigger the skill.
View source on GitHub ↗License: MIT