Skip to main content

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

ViewWhat it shows
DiscoveryProtocol version, service identity, capability summary, trust posture, endpoints
ManifestFull capability declarations — side effects, costs, scopes, inputs/outputs, signature
JWKSPublic signing keys for manifest and token verification
AuditBrowsable audit entries with filtering by capability, event class, time range
CheckpointsMerkle checkpoint list with detail inspection
InvokeForm-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:

  1. Select capability — pick from the manifest or deep-link from a capability card
  2. Auth — enter a bearer token (API key or delegation token)
  3. Permissions — auto-checks what the token allows (available / restricted / denied)
  4. Input form — generated from the manifest's input declarations with type hints and defaults
  5. Result — success with invocation_id, cost_actual, and result data, or structured failure with resolution guidance