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

# Reports

> Every BotGauge red-team run creates a report. Read the Reports list, the result counts, rules, techniques, and attempts, and what the unlock hints mean.

Every run creates a report. Open the **Reports** tab on an agent to see them all. Single-attack runs show up here too.

## The Reports list

Use the **All**, **Black box**, and **Grey box** filters to narrow the list.

| Column | Meaning |
| - | - |
| Report | The run name, for example `#_run_3` |
| Test | The test type, such as Black box |
| When | When it started and how long it took |
| Result | Found, Held, and Errors counts |
| Status | For example Completed |
| Started by | Who started the run |

## A report

The top of a report names the agent, its environment and version, when the run happened, how long it took, who started it, and how it compares with the previous run, for example "0 new breaches, 5 still failing, 1 fixed".

Seven tiles count the results of the run: **confirmed finding**, **signal**, **held**, **couldn't verify**, **not evaluable**, **blocked by BotGauge**, and **error**. See [Read your results](https://www.botgauge.com/docs/red-teaming/results.md) for what each means.

Below the tiles, each test shows how many attacks ran, how many held, and how many breached, along with how the attacks were delivered and judged, for example "Through chat, judged by the replies".

### About this run

| Section | What it tells you |
| - | - |
| Errors | How many attacks errored outside your agent, and why, for example a target timeout. These are never counted as held or as issues. |
| To test more | What you could connect to unlock more rules, such as a staging hook that lets BotGauge place context, or a read-only verifier. |
| Comparable | Whether this run can be fairly compared with others. A run can be "not comparable". |

### Attacks

The attacks section shows how many attacks ran across how many rules. You can group by rule or list all attacks, and filter by test and result, and sort by most breaches.

Each rule shows its text, whether it is **Your rule** or a **BotGauge baseline** rule, the test type, and its outcome, for example "3 of 3 reproduced", "Held, 12 attempts", or "Couldn't verify: tool not reached". A rule with a finding also shows what decided it, the evidence used, and why, with a link to **Open finding**.

Open a rule to see the **techniques** that were tried, such as Authority claim, Crescendo, Direct, Goat, and Role play. Each technique shows its number of attacks and whether any breached. Open a technique to see each attempt, marked **Held** or **Breach**.

> [!TIP]
>
> Click a number in the tiles to filter the attack list to that result.

Read every run reportReview results, techniques, and evidence for each run in [BotGauge](https://cloud.botgauge.com).

[View reports →](https://cloud.botgauge.com)

## Next

Open a finding to see the proof. See [Findings and evidence](https://www.botgauge.com/docs/red-teaming/findings.md).


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