ANIP Studio
Studio is a browser-based UI for inspecting and interacting with ANIP services. Connect to any ANIP service and browse its discovery document, manifest, keys, audit log, and checkpoints — then invoke capabilities with permissions checking and structured failure display.
Views
| View | What it shows |
|---|---|
| Discovery | Protocol version, service identity, capability summary, trust posture, endpoints |
| Manifest | Full capability declarations — side effects, costs, scopes, inputs/outputs, signature |
| JWKS | Public signing keys for manifest and token verification |
| Audit | Browsable audit entries with filtering by capability, event class, time range |
| Checkpoints | Merkle checkpoint list with detail inspection |
| Invoke | Form-based capability invocation with permissions panel and structured result/failure rendering |
Deployment modes
Embedded
Mount Studio inside your Python ANIP service at /studio:
from anip_studio import mount_anip_studio
mount_anip_studio(app, service)
# → Open http://localhost:9100/studio/
In embedded mode, Studio auto-connects to the host service. No CORS needed.
Standalone (Docker)
Run Studio as an independent container that connects to any ANIP service:
docker build -t anip-studio studio/
docker run -p 3000:80 anip-studio
# → Open http://localhost:3000, enter service URL
Standalone Studio requires the target ANIP service to have CORS enabled (see HTTP transport: CORS).
Invoke workflow
Studio's Invoke view walks through the same agent workflow the protocol defines:
- Select capability — pick from the manifest or deep-link from a capability card
- Auth — enter a bearer token (API key or delegation token)
- Permissions — auto-checks what the token allows (available / restricted / denied)
- Input form — generated from the manifest's input declarations with type hints and defaults
- Result — success with
invocation_id,cost_actual, and result data, or structured failure with resolution guidance