just-vibeGeneral · inspect by default
/trace
Follow a request, event, or data value through the system
Use to follow one execution or data path across boundaries; map shows topology without reconstructing a particular flow; data-lineage maps where a field or metric comes from across transformations.
Make it your own.
In Claude Code, use the slash command and add your context. In Codex, select trace from the just-vibe skill picker, then send the same brief.
Version 0.11.0 also supports /jv trace, /just-vibe trace and /jv:trace in Claude. See shortcut setup and context examples.
/just-vibe:trace Trace an order from the checkout route through the payment job./just-vibe:trace Trace a checkout request through a queued email and a failed payment callback./just-vibe:trace Trace only these supplied logs; downstream worker logs are unavailable.What the agent does
- Identify the entry point and start from a concrete input or correlation ID.
- Follow each transformation and dispatch, recording the hop's input, output, identity propagation and synchronous or asynchronous transition; mark missing spans where the path is unobserved.
Inputs
- request, event, identifier, or data value and optional environment/time window. Requires source or relevant sanitized telemetry.
Optional context: scope, references, constraints, successCriteria, environment, mode, budget.
Scope
- Reads
- One end-to-end execution/data path, including service boundaries; no new instrumentation by default.
- Writes
- No source changes in inspect/plan. Save only requested planning artifacts. A separately requested repair uses the relevant implementation workflow.
- Mode
- Inspect; request, event, identifier, or data value and optional environment/time window. Requires source or relevant sanitized telemetry.
- 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
- Ordered hop table with each hop's source or telemetry evidence, payload transformation, failure branches and gaps.
How the work is checked
- An async job is separated from its triggering request; an absent downstream log is reported as a gap rather than proof of success.
When to stop or clarify
- Do not scan unrelated tenant data or infer causation from timestamp proximity 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
- Locate entry point, transformations, asynchronous handoffs, identity propagation and terminal effects.
- Method
- Follow one concrete request/value, recording boundaries and error branches with file or runtime evidence.
- Pitfall
- A static call graph does not prove which branch executed; queue handoffs can change identity and ordering.
- Check
- Reconcile the trace against one success and one failure observation, marking any inaccessible runtime segment.
Situational decisions
When a queue or external service has no downstream evidence: End the confirmed trace at that boundary and list the exact log or source needed to continue.
The coding agent follows this workflow using its available tools. Installation does not grant service access or guarantee an outcome. Read the compatibility notes.