> ## Documentation Index
> Fetch the complete documentation index at: https://ito.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# PR Comments
> After each test run completes, Ito posts a comment on the GitHub pull request summarizing what was tested, what failed, and how to reproduce each bug.
When a [test run](/concepts/test-runs) finishes, Ito posts a comment directly on the GitHub pull request. The comment gives your team a full picture of the test results without leaving GitHub.
## What the comment contains
A PR comment from Ito includes:
* **Result title** — A pass (`✅`) or fail (`❌`) heading indicating the overall outcome.
* **Run summary** — The total number of test cases that ran, how many failed, and how many passed.
* **Narrative** — A plain-language description of what was tested and what the agent observed.
* **Failed test details** — For each failing test case: what failed, the impact, numbered reproduction steps, stub/mock context, and code analysis with relevant file paths and line numbers.
* **Passed test summary** — A collapsible table listing all passing test cases.
* **Additional findings** — Failures that are not attributed to the current PR's changes are surfaced separately so they do not affect the overall pass/fail result.
* **Commit link** — A link to the exact commit SHA that was tested.
* **Full run link** — A link to the detailed run view in the Ito dashboard.
Additional findings are pre-existing issues observed during the run that are unrelated to the changes in this PR. They are informational and do not cause the run to be marked as failed.
## Diff report comments
When [incremental review](#review-scope) is enabled, follow-up commits on a PR receive a **diff report** comment instead of a full report. The diff report focuses on what changed since the previous commit and clearly separates tests Ito verified on this commit from tests it carried forward from the prior run.
### Headline
The diff report headline has two lines so the math is internally consistent:
* **`X test cases ran this commit`** — The verified count, broken down into `passed ✅`, `fixed 🔧`, `failed ❌`, and `additional findings ⚠️`. These are the tests Ito actually executed against the current commit.
* **`↪️ Carried forward from prior run`** — Counts of tests that were not retested this commit because Ito judged them unaffected by the diff. Reported separately so they do not inflate the trust signal.
If any tests are no longer applicable (their target code was removed or renamed), a third line summarizes that too.
### Sections
The diff report groups results into the following collapsible sections:
| Section | What it contains |
| ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------- |
| `❌ Failures` | All current failures, both verified this commit and inherited from the prior run. Includes an **Origin** column. |
| `✅ Bugs Fixed` | Tests that were failing in the prior run and now pass. |
| `✅ Verified Passing` | Tests that Ito ran this commit and passed. |
| `↪️ Inherited from Prior Run` | Tests that passed previously and were not retested this commit because they were unaffected by the diff. |
| `⚠️ Additional Findings` | Failures unrelated to this PR's changes. Includes an **Origin** column distinguishing newly observed from inherited findings. |
| `🗑️ No Longer Applicable` | Prior tests whose target code has been removed or renamed. Retired, not failing. |
### Origin column
The `Failures` and `Additional Findings` tables include an **Origin** column so reviewers can gauge trust without expanding details. Failures are sorted top-down by actionability:
* `🆕 New` — A new test that failed on this commit.
* `🔁 Regression` — A test that passed previously and now fails.
* `🔻 Still broken (verified)` — A failure Ito retested and confirmed on this commit.
* `🔻 Still broken (inherited)` — A failure carried forward from the prior run; not retested this commit.
Additional findings use the same Origin convention with `🆕 New` for findings observed this commit and `↪️ Inherited` for carry-forward findings.
When triaging a diff report, start at the top of the `Failures` table. Newly introduced failures and regressions are sorted first; inherited carry-forward entries appear last.
## Configurable options
You can control what appears in PR comments from **Settings → Automation**. See [Automation settings](/settings/automation) for the full reference.
### Comment content
Include video recordings of the agent's session in the comment so reviewers can watch what was tested directly from GitHub.
Show an estimated amount of engineering time saved by automated testing in the PR comment.
### Review scope
The **Review Scope** settings control how Ito decides when to run and post comments:
* **Automatic Review** — Automatically run a full review on every new pull request without manual triggering. Enabled by default.
* **Automatic Incremental Review** — Run tests only for areas affected by the latest changes instead of the full application. Enabled by default.
Incremental reviews are faster because Ito focuses on the diff rather than re-testing the entire application. Full reviews are more thorough and recommended for high-risk changes.
## Automation settings
You can further control automated review behavior from [**Settings → Automation**](/settings/automation):
* **Automated Reviews** — A master switch that enables or disables all automated review activity for the repository.
* **Smart Defaults** — Automatically apply recommended settings based on your repository's size and activity.
## Related
* [Test Runs](/concepts/test-runs) — The run that triggers the comment
* [Test Cases](/concepts/test-cases) — The individual results summarized in the comment
* [Severity Levels](/concepts/severity-levels) — How failures are classified in the comment