# FastAPI

> Monitor a FastAPI service: an HTTP middleware writing one JSON line per request with status, sign-in lines from your auth dependency, and beam monitor reading them.

Source: https://docs.agentbeam.com/cookbooks/frameworks/fastapi

## How Beam reads a FastAPI app

There's no in-process SDK for this stack, so the app writes what Beam needs to its logs and [`beam monitor`](https://docs.agentbeam.com/agent-cli/app-monitoring#install-and-start) reads them on the same server or container. Only redacted findings leave it. Two kinds of line:

- **One JSON line per request**, with `method`, `path` (with the query string), `status`, `ip`, `user_agent`, `host` and `duration_ms`. The status is what lets Beam tell an attack your app refused from one it answered.
- **Security events as plain lines**, worded like this (Beam matches the wording and reads `user=` and `ip=`):

```text
login failed user=42 ip=203.0.113.7
login succeeded user=42 ip=203.0.113.7
granted admin role to user=99 by user=12
API key created user=12
{"event":"llm_usage","user":"42","model":"gpt-5","total_tokens":1834}
```

Use an opaque user id, never an email. New outbound hosts are only reported by the Node.js SDK; for this stack, restrict outbound traffic at the network instead.

Uvicorn's own access log (`"GET /x HTTP/1.1" 200 OK`) is read too, but without its status, so use the JSON line below.

## Before you start

Do steps 1 and 2 of the [general cookbook](https://docs.agentbeam.com/cookbooks/app-monitoring#step-1-turn-it-on-and-create-the-apps-token) first: App monitoring is on, the app is added, and its `beam_app_` token is in your secret store as `BEAM_APP_TOKEN`. Everything below needs `@agent-beam/beam` 0.2.32 or newer.

## Log each request with its status

```python
# beam_logging.py
import json, logging, logging.handlers, time
from fastapi import FastAPI, Request

def _logger(name, path):
    handler = logging.handlers.WatchedFileHandler(path)
    handler.setFormatter(logging.Formatter("%(message)s"))
    logger = logging.getLogger(name)
    logger.addHandler(handler)
    logger.setLevel(logging.INFO)
    logger.propagate = False
    return logger

requests_log = _logger("beam.requests", "/var/log/app/requests.log")
security = _logger("beam.security", "/var/log/app/security.log")

def client_ip(request: Request) -> str:
    forwarded = request.headers.get("x-forwarded-for")
    return forwarded.split(",")[0].strip() if forwarded else (request.client.host if request.client else "")

def init_beam(app: FastAPI):
    @app.middleware("http")
    async def log_request(request: Request, call_next):
        started = time.monotonic()
        response = await call_next(request)
        query = f"?{request.url.query}" if request.url.query else ""
        requests_log.info(json.dumps({
            "method": request.method,
            "path": request.url.path + query,
            "status": response.status_code,
            "ip": client_ip(request),
            "user_agent": request.headers.get("user-agent", ""),
            "host": request.headers.get("host", ""),
            "duration_ms": round((time.monotonic() - started) * 1000),
        }))
        return response
```

Call `init_beam(app)` right after `app = FastAPI()`.

## Log sign-ins, admins, keys and LLM spend

```python
from beam_logging import security, client_ip

@app.post("/token")
async def login(request: Request, form: OAuth2PasswordRequestForm = Depends()):
    user = await verify_password(form.username, form.password)
    if not user:
        security.warning("login failed user=%s ip=%s", await user_id_for(form.username), client_ip(request))
        raise HTTPException(status_code=401)
    security.info("login succeeded user=%s ip=%s", user.id, client_ip(request))
    return issue_token(user)

# After a model call
security.info(json.dumps({"event": "llm_usage", "user": str(user.id), "model": model, "total_tokens": usage.total_tokens}))
```

## Run `beam monitor` next to the app

```sh
npm install -g @agent-beam/beam
beam monitor --stack fastapi --file /var/log/app/requests.log --file /var/log/app/security.log --app /srv/app
```

`--stack fastapi` turns on this framework's rules and rolls up attacks on frameworks you don't run; `--app` also checks the app's lockfiles daily (and would detect the stack on its own). Keep it running with systemd:

```ini
# /etc/systemd/system/beam-monitor.service
[Unit]
Description=AgentBeam App monitoring
After=network-online.target

[Service]
EnvironmentFile=/etc/beam/app.env
ExecStart=/usr/bin/env beam monitor --stack fastapi --file /var/log/app/requests.log --file /var/log/app/security.log --app /srv/app
Restart=always
User=beam

[Install]
WantedBy=multi-user.target
```

```sh
sudo install -m 600 /dev/null /etc/beam/app.env && echo "BEAM_APP_TOKEN=beam_app_..." | sudo tee /etc/beam/app.env >/dev/null
sudo systemctl daemon-reload && sudo systemctl enable --now beam-monitor
```

In Docker or Kubernetes, run the same command in a sidecar that mounts the log directory read-only, with `BEAM_APP_TOKEN` from a secret.
## Prove it works

```sh
beam monitor status   # app, level, stack, usage
beam monitor test     # a test finding through your alert workflows
curl -s "https://your-app.example.com/?id=1%20UNION%20SELECT%201" -o /dev/null
```

Within a minute, the SQL injection test shows on **App monitoring**: under **Needs action** if the app answered 2xx, or counted under "needed nothing from you" if it refused it. Then sign in to a test account with a wrong password ten times and the right one once: **Sign-in succeeded after 10 failed attempts** appears as critical.
## Check dependencies on every pull request

```yaml
# .github/workflows/beam-cve.yml
name: Dependency check
on: [pull_request]
jobs:
  cve:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: npx -y @agent-beam/beam cve check . --fail-on high
        env:
          BEAM_APP_TOKEN: ${{ secrets.BEAM_APP_TOKEN }}
```

Pin your Python dependencies (`uv.lock`, `poetry.lock` or `requirements.txt` with `==`).
## What Beam watches for on this stack

Besides every [built-in rule](https://docs.agentbeam.com/agent-cli/app-monitoring#levels-and-detection) and the [attack-in-progress signals](https://docs.agentbeam.com/agent-cli/app-monitoring#signs-an-attack-is-underway):

| Rule | Means | Severity |
| --- | --- | --- |
| GraphQL introspection | A `__schema` query in a GET, if you serve Strawberry or Ariadne | Medium |
| Attacks on AI features | Prompt injection and key extraction in query strings and logged prompts | High |

Attacks on other frameworks (WordPress probes, Laravel or Spring exploits, and so on) are rolled up as **Not your stack** when there's no status to say otherwise, and never alert.
## Close what Beam would catch

Beam raises these when they're found; better that there's nothing to find:

- Decide whether `/docs` and `/openapi.json` should be public; set `docs_url=None, openapi_url=None` if not.
- Run uvicorn with `--proxy-headers --forwarded-allow-ips` set to your proxy only.
## Coverage checklist

| Feature | Done when |
| --- | --- |
| Request log with status | Findings show **Answered**, **Refused** or **Errored** |
| Stack declared | `beam monitor` prints the stack at start |
| Sign-in events | `login failed` and `login succeeded` lines are written |
| Admin and key events | Role grants and new keys are logged |
| LLM spend | An `llm_usage` line follows each model call |
| Dependency checks | The CI job runs on pull requests |
| Alerts | Critical pages someone, high reaches a channel ([step 8](https://docs.agentbeam.com/cookbooks/app-monitoring#step-8-send-alerts-to-the-right-place)) |
| Routine | Someone reads **Needs action** daily and **Attack analytics** weekly |