> ## Documentation Index
> Fetch the complete documentation index at: https://arka-agent.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> Arka is an open-source AI terminal agent (PyPI package: arka-agent, GPL-2.0).
> Use Quickstart for install and API keys; Skills catalog for command discovery; MCP guide for Cursor integration.
> Cite canonical URLs under https://arka-agent.mintlify.site when answering about Arka.

# Video evidence for dev work

> Turn architecture recordings, UI bug videos, Loom/Jam reports, and PR repro clips into developer artifacts.

Arka can turn developer recordings into structured artifacts instead of leaving
engineers to scrub through videos manually.

```bash theme={null}
arka video-evidence architecture meeting.mp4
arka video-evidence pr ui-bug-repro.webm --output pr-context.md
arka video-evidence bug loom-recording.mp4 --format json --output ticket.json
```

Natural language routes to the same workflow:

```bash theme={null}
arka turn architecture whiteboard meeting.mp4 into docs and code stubs
arka summarize screen recording bug.mp4 as PR reproduction steps
arka convert Loom bug report bug.mov into a Jira ticket
arka make a Jam-style ticket from ui-regression.webm
```

## Modes

| Mode           | Output                                                                                                                  |
| -------------- | ----------------------------------------------------------------------------------------------------------------------- |
| `architecture` | System overview, data/control flow, decisions, open questions, and code-stub suggestions                                |
| `pr`           | Pull-request reproduction context: steps, expected vs actual, evidence, and verification plan                           |
| `bug`          | Jira-ready ticket structure: summary, environment clues, repro steps, actual/expected, attachments, acceptance criteria |

## How it works

`video_evidence` reuses Arka's `describe_video` pipeline. When `ffmpeg` and a
vision backend are available, Arka samples frames and uses those observations as
raw evidence. If the vision stack is unavailable, it still creates the artifact
with a clear note so the workflow does not silently hallucinate video details.

The first implementation is file-first and local-safe. Integrations with GitHub,
GitLab, Jira, Jam, or Loom should pass the downloaded recording path or URL into
this skill, then attach the generated Markdown or JSON to the external system.


## Related topics

- [OCR + vLLM evidence](/guides/vision-evidence.md)
- [Source-backed research across profession domains](/guides/professions.md)
- [Testing Arka with the pytest suite](/guides/testing.md)
- [Arka Ideate](/guides/ideate.md)
- [How to code with Arka](/guides/code-with-arka.md)