vitna.ai/docs/api/v1/verification/evidence Machine view
Raw .md llms.txt
> Source: https://vitna.ai/docs/api/v1/verification/evidence
> Raw Markdown: https://vitna.ai/docs/api/v1/verification/evidence.md
> Corpus index: https://vitna.ai/llms.txt

# Evidence

The record of one archived document, and behind a token, the document itself.

```
GET /api/v1/evidence/{sha256}
```

```bash
curl https://vitna.ai/api/v1/evidence/9f2c...
curl -H 'Authorization: Bearer <token>' "https://vitna.ai/api/v1/evidence/9f2c...?body=1"
```

## Path parameters

| Parameter | Meaning |
|---|---|
| `sha256` | 64 lowercase hex characters |

Anything else answers `404 unknown_evidence` without touching the archive.

## The public answer

```json
{
  "sha256": "9f2c...",
  "archived": {
    "url": "https://...", "fetchedAt": "...", "httpStatus": 200,
    "headers": { ... }, "userAgent": "...", "adapter": "...",
    "parserVersion": 3, "bytes": 48211, "via": null, "robots": null
  },
  "reads": [ ... ],
  "body": { "available": true, "url": "/api/v1/evidence/9f2c...?body=1", "access": "Bearer token. ..." },
  "meta": { "generatedAt": "...", "source": "archive" }
}
```

`archived` is the read that first stored this document. `reads` is every fetch
that has since found the same bytes, which is how you see that a page has been
unchanged for weeks.

## Why the body is gated

The public half is what a verification needs. Given the URL and the fetch time,
you can fetch the page yourself and compare hashes; Vitna does not need to be
trusted for that to work.

The stored copy is a third party's page, and republishing it raises a copyright
question this record does not need to answer in order to be checkable.

## The body response

Not JSON. The stored bytes, with the archived `content-type`, plus:

| Header | Meaning |
|---|---|
| `x-vitna-fetched-at` | When it was read |
| `x-vitna-source-url` | What was read |
| `x-vitna-sha256` | The key |

It is sent with `no-transform`, so intermediaries such as Cloudflare pass the
bytes through unaltered and the body hashes to the key.

## Errors

| Status | Code | Cause |
|---|---|---|
| 404 | `unknown_evidence` | Malformed hash, or nothing archived under it |
| 401 | `unauthorized` | `?body=1` without a valid bearer token |

The `401` message restates that the hash, URL and fetch time are public and are
what a verification needs, so a caller without a token knows they are not
blocked from checking anything.

## Caching

`no-store`. Both the record and the body; the body adds `no-transform`.
Human Machine Shift+M