> ## Documentation Index
> Fetch the complete documentation index at: https://parser.xdr.ooo/llms.txt
> Use this file to discover all available pages before exploring further.

# Scenario Playbooks: choose what happened

> Find the next step based on what Auto Parser or ChatGPT showed, not the input format.

# Choose what happened after analyzing your logs

**Everyone starts in Auto Parser:** paste a few complete source events and click **Analyze Logs**. You do not have to identify a log format.

## Auto Parser: solve it inside the guided workflow

<CardGroup cols={2}>
  <Card title="All logs were recognized" icon="circle-check" href="/scenarios/auto-success">
    Continue to Review Fields, confirm suggested mappings, then create and download the files.
  </Card>

  <Card title="Part of an event is missing" icon="triangle-exclamation" href="/scenarios/partial-event">
    Select Problem to review, inspect the highlighted region, and try the recommended fix.
  </Card>

  <Card title="My product sends several event types" icon="layers" href="/scenarios/mixed-events">
    Paste examples together, review every type, and create one combined draft only when supported.
  </Card>

  <Card title="Values have no reliable field names" icon="table" href="/scenarios/unnamed-values">
    Keep all values with temporary names or enter trusted names and confirm the column order.
  </Card>

  <Card title="A field has the wrong output name" icon="arrow-left-right" href="/scenarios/normalization">
    Correct its destination and type under Map Fields instead of rewriting extraction.
  </Card>

  <Card title="One event spans several lines" icon="align-left" href="/scenarios/long-events">
    Review the message boundary, then continue only with a supported framing rule.
  </Card>

  <Card title="I want to save parser versions" icon="history" href="/scenarios/revisions">
    Compare experiments and restore a browser-local draft.
  </Card>

  <Card title="I need changing sample logs" icon="file-code" href="/scenarios/test-sender">
    Download and use the standalone Python test sender.
  </Card>
</CardGroup>

## ChatGPT Assisted Parser: recover or verify a difficult case

Use the left-side **ChatGPT Assisted Parser** menu when the automatic path cannot safely interpret the source or proposes a change that needs independent checking.

<CardGroup cols={2}>
  <Card title="Nothing in my log was understood" icon="comments" href="/scenarios/unknown-log">
    Create a manually reviewed, masked prompt and paste the restricted JSON result back.
  </Card>

  <Card title="One missing field needs a new rule" icon="wrench" href="/scenarios/recipe-ai">
    Preserve correct values and request one narrow change.
  </Card>

  <Card title="ChatGPT's answer was rejected" icon="circle-xmark" href="/scenarios/ai-invalid-response">
    Read the error and ask for a supported structured correction.
  </Card>

  <Card title="The suggested fix needs additional proof" icon="check-double" href="/scenarios/ai-test-change">
    Use other events as independent tests before approving the result.
  </Card>
</CardGroup>

## A real screenshot of a flagged event

<img src="https://mintcdn.com/xdr-parser/DkeLTskQMJz9hCfD/images/07-unknown-analysis.png?fit=max&auto=format&n=DkeLTskQMJz9hCfD&q=85&s=ca5b0d595da9ea9926950e92b8ca067c" alt="Parser Studio highlighting unresolved source text and showing the Fix with ChatGPT action" width="1440" height="2175" data-path="images/07-unknown-analysis.png" />

*If any events need review, the application highlights the source section instead of silently discarding it.*

## Which guide should I read first?

* [Everything parsed correctly → finish the four steps](/quickstart)
* [Some text is highlighted → repair only the affected part](/scenarios/partial-event)
* [Completely unfamiliar logs → ChatGPT Assisted Parser](/scenarios/unknown-log)
* [One product, many event types → review a combined parser](/scenarios/mixed-events)

All images were captured from the real Parser Studio interface using fictional data. No extra parser mode is required.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.