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

# Thread reviews

> Inspect what CodeRabbit Agent for Slack did in a Slack thread with run history, context, diffs, and execution details.

**See exactly what happened and why.** Thread reviews give you a web-based inspection surface for any CodeRabbit Agent for Slack conversation, from the high-level outcome down to individual commands, diffs, and execution detail. Use them when you need more than the final Slack reply.

## How to open a thread review

Teams typically reach a thread review from:

* A link in [Usage](/slack-agent/usage)
* A direct thread review URL
* Links CodeRabbit Agent provides in a specific run or thread, when available

Slack-backed authorization still applies. If the viewer does not have access to the underlying Slack thread, the review cannot be opened.

Thread review access is also tied to the same Slack workspace as the original thread. For private channels and non-bot DMs, the web app may require Slack authorization to verify that the viewer can still see the conversation. A user can still reopen reviews for their own 1:1 bot DM without that extra visibility check.

## From overview to execution detail

Thread reviews are organized in three layers:

1. **Thread overview**: the high-level summary covering all runs in the thread
2. **Individual runs**: each run's outcome, timing, context usage, and status
3. **Execution detail**: the deepest layer showing commands, tool calls, and moment-by-moment activity

## Two inspection tabs

Thread reviews provide two main views:

<Tabs>
  <Tab title="Conversation">
    The **Conversation** tab shows the narrative flow of the run:

    * The execution hero (goal, outcome, draft status)
    * Key moments such as investigation, editing, tool use, and recovery
    * Proof sections linking to the commands and changes that produced the result
    * Why the result should be trusted (the narrative "review story")
  </Tab>

  <Tab title="Diff">
    The **Diff** tab shows the code changes produced by the run:

    * **Draft** diff: current live or unsaved work
    * **Run** diff: changes from a single selected run
    * **All** diff: combined history of changes across the entire thread

    Each mode shows repo-level and per-file patches with direct links to changed files.
  </Tab>
</Tabs>

## What you can inspect

Across both tabs, common review layers include:

* Run history and status across the thread
* Context used during each run (repositories, connections, knowledge)
* Diffs and changed files with per-file patches
* Narrative moments (investigate, edit, tool use, recovery, reply)
* Exported artifacts and deep links

## Navigating between runs

When a thread has multiple runs, the thread overview lists them in order. Select any run to see its execution hero, conversation narrative, and diffs. Earlier runs are compressed so you can focus on the most recent result while still having full history available.

## Best way to read a thread review

1. Start with the thread overview and execution hero.
2. Open the most relevant run.
3. Check the Conversation tab for the narrative and proof.
4. Switch to the Diff tab when you need to see exactly what changed.
5. Drop into deeper execution layers only when you need to understand *how* the result was produced.

## What's next

<CardGroup cols={1}>
  <Card title="Usage" href="/slack-agent/usage" icon="bar-chart-3" horizontal>
    Find the run you care about and jump into its review surface.
  </Card>

  <Card title="Knowledge Base" href="/slack-agent/knowledge-base" icon="brain" horizontal>
    Preserve useful facts after reviewing what actually happened in the thread.
  </Card>

  <Card title="Plans" href="/slack-agent/plans" icon="list-checks" horizontal>
    Review how planning work created in Slack is preserved, compared, and handed off.
  </Card>
</CardGroup>