Agent Manifest

Registration Demo — Agent Manifest

This document demonstrates the full operational registration pipeline using the Agent Manifest infrastructure.

The example below follows a real registration flow.

---

1. Generate the Manifest

The process begins using the Agent Manifest Ambassador.

👉 https://agent-manifest.github.io/agent-manifest-ambassador/

The user provides the required declaration fields and the Ambassador produces a manifest JSON.

Example declaration:

{
  "$schema": "https://agent-manifest-spec.org/spec/v1.0/schema.json",
  "manifest_version": "1.0",
  "agent_id": "agent-manifest-ambassador",
  "agent_name": "Agent Manifest Ambassador",
  "agent_version": "0.1.0",

  "owner": {
    "type": "organization",
    "identifier": "agent-manifest"
  },

  "purpose": {
    "primary_code": "general.assistance",
    "description": "Conversational generator that produces valid Agent Manifest declarations."
  },

  "forbidden_actions": [
    "execute_external_actions",
    "store_user_data_beyond_session"
  ],

  "autonomy": {
    "level": 1
  },

  "risk_profile": {
    "level": "low"
  },

  "data_handling": {
    "stores_personal_data": false
  },

  "stopping_authority": {
    "stoppable_by": ["owner"],
    "mechanism": "Manual disable via hosting interface."
  },

  "audit_surface": {
    "logging": "basic",
    "reconstructability": "partial"
  },

  "contact": {
    "email": "contact@agent-manifest-spec.org"
  }
}

Full structural reference: spec/v1.0/spec.md

---

2. Create the Submission

After generating the manifest, the Ambassador opens a pre-filled GitHub Issue in the dataset repository:

👉 https://github.com/agent-manifest/agent-manifest-dataset

The issue includes:

• the manifest JSON
• the agent identity
• submission metadata

The user reviews the content and submits the issue.

---

3. Workflow Execution

Submitting the issue triggers an automated workflow.

The workflow performs:

1. JSON extraction from the issue
2. schema validation
3. field verification
4. preparation of the dataset entry

If the manifest structure is invalid, the workflow reports an error.

---

4. Manifest Stored in Dataset

If validation succeeds, the manifest is stored in the dataset repository.

Storage path format:

manifests/YYYY/MM/agent-name.json

Example:

manifests/2026/03/agent-manifest-ambassador.json

This creates a permanent public declaration record.

---

5. Registry Update

After the manifest is stored, the registry index is updated.

The registry provides a machine-readable list of registered agents.

Each entry references the stored manifest.

---

6. Issue Closed Automatically

If the registration process succeeds:

• the workflow posts a confirmation comment
• the issue is automatically closed

This provides a transparent public trace of the registration.

---

7. Discovery Endpoint

External systems can discover registered agents through the well-known endpoint:

https://agent-manifest-spec.org/.well-known/agent-manifest-registry.json

This endpoint exposes the registry in machine-readable form.

---

Result

The full registration pipeline produces:

Manifest JSON
↓
GitHub Issue submission
↓
Automated workflow validation
↓
Manifest stored in dataset
↓
Registry index updated
↓
Discovery endpoint available

This demonstrates the operational infrastructure of the Agent Manifest system.

This site is open source. Improve this page.