Skip to content
Back to skills

Frappe Syntax Whitelisted

ASecurity

Use when creating Frappe Whitelisted Methods (Python API endpoints) for v14/v15/v16. Covers @frappe.whitelist() decorator, frappe.call/frm.call invocations, permission checks, error handling, response formats, and client-server communication. Keywords: whitelisted, API endpoint, frappe.call, frm.call, REST API, @frappe.whitelist, allow_guest, API endpoint example, frappe.whitelist syntax, how to expose function.

  • 174 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added May 27, 2026
data-aijavascriptpythonjavabashsqlapisecurityperformance

Works with

  • claude code
  • cli
  • api

Security analysis

A100/100

Pro scans all 13 files and shows the line behind each finding

Scanned May 27, 2026

npx -y skills add Impertio-Studio/Frappe_Claude_Skill_Package --skill frappe-syntax-whitelisted --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Frappe Syntax Whitelisted?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Frappe Syntax Whitelisted
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/impertio-studio-frappe-syntax-whitelisted/badge)](https://www.skillsdirectory.com/skills/impertio-studio-frappe-syntax-whitelisted)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: frappe-syntax-whitelisted
description: >
  Use when creating Frappe Whitelisted Methods (Python API endpoints) for
  v14/v15/v16. Covers @frappe.whitelist() decorator, frappe.call/frm.call
  invocations, permission checks, error handling, response formats, and
  client-server communication. Keywords: whitelisted, API endpoint,
  frappe.call, frm.call, REST API, @frappe.whitelist, allow_guest,
  API endpoint example, frappe.whitelist syntax, how to expose function.
license: MIT
compatibility: "Claude Code, Claude.ai Projects, Claude API. Frappe v14-v16."
metadata:
  author: OpenAEC-Foundation
  version: "2.0"
---
# Frappe Syntax: Whitelisted Methods

Whitelisted methods expose Python functions as HTTP API endpoints via `/api/method/`.

## Quick Reference

```python
import frappe
from frappe import _

# Authenticated endpoint (default)
@frappe.whitelist()
def get_customer_summary(customer):
    frappe.has_permission("Customer", "read", throw=True)
    return frappe.get_doc("Customer", customer).as_dict()

# Public endpoint — ALWAYS validate input thoroughly
@frappe.whitelist(allow_guest=True, methods=["POST"])
def submit_contact(name, email, message):
    if not name or not email:
        frappe.throw(_("Name and email required"), frappe.ValidationError)
    return {"success": True}

# Controller method — called via frm.call('method_name')
class SalesOrder(Document):
    @frappe.whitelist()
    def calculate_taxes(self, include_shipping=False):
        return {"tax": self.grand_total * 0.21}
```

**Endpoint URL**: `/api/method/myapp.module.function_name`

---

## Decorator Signature [v14+]

```python
@frappe.whitelist(
    allow_guest=False,   # True = accessible without login
    xss_safe=False,      # True = do NOT escape HTML in response
    methods=None,        # ["GET"], ["POST"], or ["GET","POST"] — default: all
    force_types=None     # True = require type annotations [v15+]
)
```

| Parameter | Default | Effect |
|-----------|---------|--------|
| `allow_guest` | `False` | `True` = Guest role can call; ALWAYS add extra input validation |
| `xss_safe` | `False` | `True` = HTML not escaped; NEVER use without sanitized output |
| `methods` | `None` (all) | Restrict allowed HTTP verbs |
| `force_types` | `None` | `True` = all params MUST have type annotations [v15+] |

Full details: [decorator-options.md](references/decorator-options.md)

---

## Decision Tree

```
What kind of endpoint?
|
+-- Standalone API (utility, integration, dashboard)?
|   --> @frappe.whitelist() on a module-level function
|   --> Call via: frappe.call('myapp.api.function')
|   --> URL: /api/method/myapp.api.function
|
+-- Document-specific action?
|   --> @frappe.whitelist() on a Document class method
|   --> Call via: frm.call('method_name')
|   --> URL: /api/method/run_doc_method (internal)
|
+-- Server Script (no-code)?
    --> Use Server Script DocType instead (no decorator needed)

Who may call the API?
|
+-- Anyone (including guests)?
|   --> allow_guest=True + thorough input validation + rate limiting
|
+-- Logged-in users only?
    +-- Specific role? --> frappe.only_for("RoleName")
    +-- DocType-level? --> frappe.has_permission(doctype, ptype, throw=True)
    +-- Document-level? --> frappe.has_permission(doctype, ptype, doc, throw=True)

Which HTTP methods?
|
+-- Read only? --> methods=["GET"]
+-- Write only? --> methods=["POST"]
+-- Both? --> methods=["GET","POST"] or default
```

---

## Permission Patterns

ALWAYS check permissions inside every whitelisted method. The `@frappe.whitelist()` decorator only verifies the user is logged in — it does NOT check DocType or document-level permissions.

```python
# DocType-level permission (throw=True raises PermissionError automatically)
@frappe.whitelist()
def get_orders():
    frappe.has_permission("Sales Order", "read", throw=True)
    return frappe.get_all("Sales Order", limit=20)

# Document-level permission
@frappe.whitelist()
def get_order(name):
    frappe.has_permission("Sales Order", "read", name, throw=True)
    return frappe.get_doc("Sales Order", name).as_dict()

# Role-based restriction
@frappe.whitelist()
def admin_action():
    frappe.only_for("System Manager")  # throws if user lacks role
    return {"secret": "data"}
```

Full patterns: [permission-patterns.md](references/permission-patterns.md)

---

## Parameter Handling

Parameters arrive as **strings** from HTTP requests. ALWAYS convert explicitly.

```python
@frappe.whitelist()
def calculate(amount, quantity, items=None):
    amount = float(amount)          # ALWAYS cast numeric params
    quantity = int(quantity)
    if isinstance(items, str):      # ALWAYS parse JSON strings
        items = frappe.parse_json(items)
    return amount * quantity
```

Access all request parameters via `frappe.form_dict`:
```python
@frappe.whitelist()
def dynamic_handler():
    all_params = frappe.form_dict
    customer = frappe.form_dict.get("customer")
```

### Type Annotations [v15+]

Frappe v15+ validates type annotations automatically at request time via Pydantic:

```python
@frappe.whitelist()
def get_orders(customer: str, limit: int = 10, active: bool = True) -> dict:
    # Frappe auto-validates: limit MUST be convertible to int
    return {"orders": frappe.get_all("Sales Order", limit=limit)}
```

### force_types and require_type_annotated_api_methods [v15+]

- `@frappe.whitelist(force_types=True)` — EVERY parameter MUST have a type annotation
- App-level enforcement via `hooks.py`: `require_type_annotated_api_methods = 1`
- Missing annotations raise `FrappeTypeError`

Full details: [parameter-handling.md](references/parameter-handling.md)

---

## Client Calls

### frappe.call(): Standalone APIs

```javascript
// Promise-based (ALWAYS prefer this)
frappe.call({
    method: 'myapp.api.get_summary',
    args: { customer: 'CUST-001' },
    freeze: true,
    freeze_message: __('Loading...')
}).then(r => {
    console.log(r.message);  // return value is in r.message
}).catch(err => {
    frappe.show_alert({ message: __('Error'), indicator: 'red' });
});
```

### frm.call(): Controller Methods

```javascript
frm.call('calculate_taxes', { include_shipping: true })
    .then(r => frm.set_value('tax_amount', r.message.tax_amount));
```

### REST API (External Clients)

```bash
# Token auth (ALWAYS use for external integrations)
curl -H "Authorization: token api_key:api_secret" \
     -H "Content-Type: application/json" \
     -X POST https://site.com/api/method/myapp.api.create_order \
     -d '{"customer": "CUST-001"}'
```

Full patterns: [client-calls.md](references/client-calls.md)

---

## Error Handling

```python
@frappe.whitelist()
def process_order(order_id):
    if not order_id:
        frappe.throw(_("Order ID required"), frappe.ValidationError)

    if not frappe.has_permission("Sales Order", "write", order_id):
        frappe.throw(_("Not permitted"), frappe.PermissionError)

    try:
        result = heavy_operation(order_id)
        return {"success": True, "data": result}
    except Exception:
        frappe.log_error(frappe.get_traceback(), "process_order")
        frappe.throw(_("Operation failed. Contact support."))
```

| Exception | HTTP Code | When to Use |
|-----------|-----------|-------------|
| `frappe.ValidationError` | 417 | Input validation failure |
| `frappe.PermissionError` | 403 | Access denied |
| `frappe.DoesNotExistError` | 404 | Document not found |
| `frappe.DuplicateEntryError` | 409 | Duplicate record |
| `frappe.AuthenticationError` | 401 | Not logged in |

Full patterns: [error-handling.md](references/error-handling.md)

---

## Response Patterns

```python
# Return value auto-wraps as {"message": <return_value>}
@frappe.whitelist()
def get_data():
    return {"key": "value"}   # Client receives: {"message": {"key": "value"}}

# Custom HTTP status
@frappe.whitelist()
def create_item(data):
    doc = frappe.get_doc(data).insert()
    frappe.local.response["http_status_code"] = 201
    return {"name": doc.name}

# File download
@frappe.whitelist()
def download_report(name):
    content = generate_pdf(name)
    frappe.response.filename = f"{name}.pdf"
    frappe.response.filecontent = content
    frappe.response.type = "download"
```

Full patterns: [response-patterns.md](references/response-patterns.md)

---

## Rate Limiting [v14+]

```python
from frappe.rate_limiter import rate_limit

@frappe.whitelist(allow_guest=True)
@rate_limit(limit=5, seconds=60)  # 5 requests per 60 seconds per IP
def public_endpoint():
    return {"status": "ok"}
```

`rate_limit` signature:
```python
rate_limit(key=None, limit=5, seconds=86400, methods="ALL", ip_based=True)
```

ALWAYS apply `@rate_limit` on `allow_guest=True` endpoints to prevent abuse.

---

## Version Differences

| Feature | v14 | v15+ | v16+ |
|---------|-----|------|------|
| `@frappe.whitelist()` | Yes | Yes | Yes |
| `allow_guest`, `xss_safe`, `methods` | Yes | Yes | Yes |
| Type annotation validation | No | Yes (auto via Pydantic) | Yes |
| `force_types` parameter | No | Yes | Yes |
| `require_type_annotated_api_methods` hook | No | Yes | Yes |
| `@rate_limit()` decorator | Yes | Yes | Yes |
| `FrappeTypeError` for missing annotations | No | Yes | Yes |

---

## Critical Rules

1. **NEVER skip permission checks** — `@frappe.whitelist()` only confirms login, not authorization
2. **NEVER use user input in raw SQL** — ALWAYS use parameterized queries or ORM
3. **NEVER leak stack traces** — log with `frappe.log_error()`, show generic messages
4. **ALWAYS validate input types** — parameters arrive as strings from HTTP
5. **ALWAYS apply `@rate_limit` on guest endpoints** — prevents abuse
6. **NEVER use `ignore_permissions=True` without a preceding role check**
7. **ALWAYS use `JSON.stringify()` for complex JS args** — arrays and objects

Full anti-patterns: [anti-patterns.md](references/anti-patterns.md)

---

## Security Checklist

For EVERY whitelisted method, verify:

- [ ] Permission check present (`frappe.has_permission()` or `frappe.only_for()`)
- [ ] Input validated (types, ranges, formats)
- [ ] SQL queries parameterized (NEVER string interpolation)
- [ ] Error messages contain no internal details
- [ ] `allow_guest=True` only with explicit reason + rate limiting
- [ ] `ignore_permissions=True` only with preceding role check
- [ ] HTTP methods restricted where possible
- [ ] Response contains only necessary fields (no sensitive data leaks)

---

## Reference Files

| File | Content |
|------|---------|
| [decorator-options.md](references/decorator-options.md) | All `@frappe.whitelist()` parameters and `force_types` |
| [parameter-handling.md](references/parameter-handling.md) | Request parameters, type coercion, `frappe.form_dict` |
| [response-patterns.md](references/response-patterns.md) | Return types, file downloads, streaming, HTTP status |
| [client-calls.md](references/client-calls.md) | `frappe.call()`, `frm.call()`, REST API, fetch patterns |
| [permission-patterns.md](references/permission-patterns.md) | Permission checks, role guards, custom logic |
| [error-handling.md](references/error-handling.md) | Exception types, `frappe.throw()`, logging |
| [examples.md](references/examples.md) | Complete working API examples |
| [anti-patterns.md](references/anti-patterns.md) | Security mistakes and performance pitfalls |
| [hooks.md](references/hooks.md) | Declaring whitelisted methods in `hooks.py` |
| [syntax.md](references/syntax.md) | Core decorator syntax and registration mechanics |

Files in this skill

  • SKILL.md11.3 KB
  • references/anti-patterns.md8.6 KB
  • references/client-calls.md8.4 KB
  • references/decorator-options.md5.1 KB
  • references/error-handling.md6.5 KB
  • references/examples.md11.6 KB
  • references/hooks.md4 KB
  • references/methods.md4.8 KB
  • references/parameter-handling.md5.7 KB
  • references/patterns.md6.9 KB
  • references/permission-patterns.md6.3 KB
  • references/response-patterns.md5.6 KB
  • references/syntax.md3.4 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.