E.V.I. (Executable Visual Intelligence)
E.V.I. (Executable Visual Intelligence) is the OpenCharts pipeline runtime — turn any flowchart into a live AI workflow where Theo executes each node in real time, calls APIs, and produces real deliverables.
In depth
Most flowcharts are static documentation; E.V.I. makes them executable. Every node maps to a Skill — a concrete behavior like “analyze this data,” “send an email,” or “create a presentation.” Click Run with Theo and Theo walks each node in order, passing data along, lighting up the canvas as it goes.
E.V.I. is also an open protocol: flows are exportable as `.evi.json` and runnable by any compatible runtime. The pre-flight clarification pass collects any missing inputs before execution, and Human Gates pause the flow for approvals.
The user’s “EDI” shorthand for this concept and the official E.V.I. acronym refer to the same system.
Examples
- Lead-intake → AI triage → Slack handoff
- RFP draft → human approval → email send
Also known as
Want to put this concept to work in OpenCharts?
Run a flow with E.V.I.Related terms
Execution Canvas
The Execution Canvas is the live view at /editor/[id]?mode=live — a flowchart editor that lights up in real time as Theo runs each node, with logs, outputs, and an inspect panel.
Skill
A skill is an installable extension that gives Theo a new behavior — a domain prompt, custom tools, allowed base tools, and an optional trigger (manual, event, schedule, or keyword).
Human Gate
A human gate is a node in an E.V.I. flow that pauses execution and waits for an explicit accept/reject from a human — usually delivered as a Telegram or email approval prompt.
.evi.json
`.evi.json` is the open file format for E.V.I. flows — exportable from any flowchart and re-runnable on any compatible runtime.