process-flow
Skill1 starsChanged 18 days agoMITView on GitHub
Use for operational step sequences, rework loops, handoffs, and optional swim lanes that show who owns each action.
In this repository7 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- architecture-diagramSkill
- data-flowSkill
- decision-flowSkill
- architecture-diagramsSkill
The file
Files
Process flow
Use this skill for ordered work steps. Give each action a process shape, each condition a decision shape, and each entry and exit a start or end shape. Add io for input or output and document for an artifact. Use decision-flow when the main purpose is separate approval outcomes.
Spec rules
- Each node has a unique string
idand ashape:start,end,process,decision,io, ordocument. - An edge uses
sourceandtarget; loop backs may point to an earlier node. Label decision exits. - Optional
lanesis an ordered array of{id,label}. Set a node'slaneto one of those IDs.lane_orientationishorizontalorvertical; nodes without a lane go to a trailing band. - In lane mode, omit explicit
xandy. Usetierwhen a step needs a particular flow rank.
{"title":"Approval work","direction":"TB","nodes":[{"id":"start","shape":"start"},{"id":"draft","shape":"process"},{"id":"check","shape":"decision","label":"Complete?"},{"id":"end","shape":"end"}],"edges":[{"source":"start","target":"draft"},{"source":"draft","target":"check"},{"source":"check","target":"draft","label":"Revise"},{"source":"check","target":"end","label":"Submit"}]}Render and inspect
On first use run python -m tools.bootstrap. Run python -m archdiagram.cli validate spec.json, then python -m archdiagram.cli render spec.json -f svg -o out/process.svg. Use drawio, html, pdf, pptx, or vsdx for other outputs; PDF needs python -m tools.bootstrap --pdf. Open the result and inspect the loop, rank order, and lane headers. Revise and render again. Via MCP, use get_diagram_template("process_flow"), validate_spec, and generate_diagram.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

