/explain

Make it your own.

In Claude Code, use the slash command and add your context. In Codex, select explain from the just-vibe skill picker, then send the same brief.

Version 0.11.0 also supports /jv explain, /just-vibe explain and /jv:explain in Claude. See shortcut setup and context examples.

What the agent does

1. Locate the definition and a real caller, and inspect the branches that matter for the question.
2. Walk one concrete input through transformations, outputs, side effects and failure handling with file references, at the depth and in the terminology the brief asks for; distinguish what the source shows statically from observed runtime behavior.

Inputs

• symbol, file, behavior, or question plus desired depth. Requires relevant source access.

Optional context: scope, references, constraints, successCriteria, environment, mode, budget.

Scope

Reads

Explain the selected behavior and its necessary dependencies; no refactoring or unsolicited repository-wide tutorial.

Writes

No source changes in inspect/plan. Save only requested planning artifacts. A separately requested repair uses the relevant implementation workflow.

Mode

Inspect; symbol, file, behavior, or question plus desired depth. Requires relevant source access.

Prerequisites

Resolve the user brief and inspect the relevant project or supplied evidence. External capabilities are optional unless the selected action actually needs them.

Expected output

• A causal explanation that walks one concrete input to its output with cited source locations, the edge path that matters, and unresolved runtime assumptions.

How the work is checked

• Explains both the normal path and a meaningful edge path; missing runtime configuration remains explicitly unknown.

When to stop or clarify

• Request a target only when multiple interpretations materially change the answer. Do not invent business intent from implementation alone.

Handling missing context

Infer

Resolve the named files, existing scripts, current task and earlier corrections from the conversation and repository.

Assume

Use the narrowest interpretation that completes a reversible local task; state a consequential assumption once.

Ask

Ask when competing targets or incompatible success conditions would change the result; continue independent inspection first.

Technical guidance

Evidence

Read the target definition, at least one caller, data shapes and relevant error handling.

Method

Walk one concrete input through state changes, output and side effects at the requested depth.

Pitfall

Plausible business intent cannot be inferred solely from a function name; configuration-dependent behavior remains conditional.

Check

Reconcile the walkthrough with source branches and show an edge path that changes the outcome.

Situational decisions

When explanation depends on configuration or external behavior not supplied: Separate the source-established path from conditional behavior and name the missing evidence.

The coding agent follows this workflow using its available tools. Installation does not grant service access or guarantee an outcome. Read the compatibility notes.

Keep exploring