---
title: Use attack info
description: Find which flags are in play, which team holds them and the handle each one sits behind.
---

A flag sits behind something public: a username, a record ID, a file path. **Attack info** tells you those handles for every flag still worth stealing, so your exploit can go straight to the flag instead of scanning a whole database.

## Read it on the site

Open **Attack info**. For each service it lists, per team number and flag store, the flag IDs placed in recent ticks, and shows when the list expires. A new list appears at the start of every tick.

## Fetch it from your farm

The same document is served as JSON, with no sign-in, at:

```
https://<event site>/api/attack-info
```

Fetch it at the start of each tick and hand each exploit the flag IDs for its service and target team:

```python
import requests

info = requests.get("https://ctf.example.com/api/attack-info", timeout=5).json()
for team in info["teams"]:
    for store, flag_ids in info["flag_ids"].get("notes", {}).get(str(team), {}).items():
        for flag_id in flag_ids:
            exploit(team, flag_id)
```

The format follows the ctf-gameserver `teams.json` convention, so tools written for it usually read this document as is. See [Attack info format](/players/attack-defence/reference/attack-info-format) for every field.

## What it doesn't contain

Attack info never lists flags placed in the current tick, only those from earlier ticks that are still valid. It holds no flags, only the handles they sit behind. Everyone gets the same document; it's useless without an exploit.
