---
title: "Import Agent Unified"
method: POST
path: "/v1/agent-configurations/import"
tags: ["Agent Environment", "agent-import"]
---

# Import Agent Unified

`POST /v1/agent-configurations/import`

Import an agent using the unified import endpoint.

Supports three input methods with deterministic access levels:
- agent_url:   Remote black_box - Just an endpoint URL to call
- agent_card:  Local grey_box - Upload agent configuration/card JSON
- source_code: Local white_box - Upload agent source code files

Access level is determined by input method (not content).

## Request body

- UnifiedImportRequest — Unified request model for agent import. Supports five input methods (exactly one required), each with deterministic access level: | Input Method | Mode | Access Level | Description | |--------------|--------|--------------|--------------------------------------| | agent_url | Remote | black_box | Just an endpoint URL to call | | agent_card | Local | grey_box | Upload agent configuration/card JSON | | source_code | Local | white_box | Upload agent source code files |
  - `agent_url` string, nullable — Remote black_box: Agent endpoint URL to call (e.g., https://api.example.com/agent)
  - `agent_card` object, nullable — Local grey_box: Upload agent configuration/card JSON directly
  - `source_code` SourceCodeUpload — Model for direct source code upload. Supports single file or multi-file upload for white box agent import.
    - `files` object, required — Map of filename to content. Use 'main.py' for single file uploads.
    - `entry_point` string — Main entry point file (must exist in files)
    - `framework` string — Framework identifier (for informational purposes only)
    - `agent_name` string, nullable — Agent name (extracted from code if not provided)
    - `model_name` string, nullable — Model name used by the agent
    - `description` string, nullable — Agent description
    - `system_prompt` string, nullable — System prompt if known
    - `tools` string[], nullable — List of tools/functions the agent uses
  - `framework` string, nullable — Framework identifier (for informational purposes only)
  - `entry_point` string, nullable — For white box: path to main file
  - `override_name` string, nullable — Override agent name
  - `rate_limit_requests_per_minute` integer
  - `api_key` string, nullable — API key for the agent
  - `external_id` string, nullable — External identity ID
  - `identity_provider` string, nullable — Identity provider name
  - `idp_metadata` object, nullable — IdP-specific metadata

## Response `201`

Successful Response

- UnifiedImportResponse — Response model for unified import.
  - `agent` Agent, required — Agent domain model.
    - `id` string, uuid, required
    - `agent_name` string, required
    - `model_name` string, required
    - `agent_url` string, nullable
    - `hub` string
    - `api_key` string, nullable
    - `agent_system_prompt` string, nullable
    - `rate_limit_requests_per_minute` integer — Rate limit for API requests per minute
    - `team_id` string, uuid, required
    - `created_by` string, uuid, required
    - `status` 'draft' | 'creating' | 'active' | 'under_review' | 'deprecated' | 'archived' | 'pending' | 'rejected' — Agent status values.
    - `created_at` integer, required
    - `updated_at` integer, required
    - `access_level` 'black_box' | 'grey_box' | 'white_box' — Access level for agent analysis. Black Box: URL only - behavioral testing only Grey Box: Agent configuration - structure/capabilities analysis White Box: Source code access - full code + sandbox runtime analysis
    - `agent_params` AgentParams — Model parameters for agent configuration.
      - `temperature` number, nullable
      - `top_p` number, nullable
      - `max_tokens` integer, nullable
      - `timeout` integer, nullable
      - `rag_params` object, nullable
      - `function_route_params` object, nullable
    - `mcp_config` MCPConfig — MCP (Model Context Protocol) proxy configuration.
      - `proxy_url` string, uri, nullable
      - `access_level` 'read' | 'write' | 'admin' — MCP (Model Context Protocol) access level values.
      - `server_urls` string[], nullable
    - `delegated_agents` DelegatedAgent[]
      - `url` string, required
      - `agent_card_url` string, nullable
      - `name` string, nullable
      - `capabilities` string[], nullable
    - `white_box_config` WhiteBoxConfigOutput — Source code access configuration for white box analysis.
      - `source_code` string, nullable
      - `repo_url` string, nullable
      - `github_username` string, nullable
      - `repo_branch` string, nullable
      - `entry_point` string, nullable
    - `agent_metadata` AgentMetadata — Agent metadata for display and organization.
      - `role` string, nullable
      - `job_description` string, nullable
      - `version` string, nullable
      - `author` string, nullable
      - `tags` string[], nullable
    - `purpose` string, nullable
    - `protocol` 'chat_completions' | 'a2a' — Communication protocol for agent endpoints.
    - `test_score` string, nullable
    - `last_tested_by` string, nullable
    - `evaluated_harnesses` string[], nullable
    - `last_evaluated_at` integer, nullable
    - `protection_status` 'unprotected' | 'configured' | 'domed' — Dome protection lifecycle state.
    - `trust_stage` 'registered' | 'tested' | 'hardened' | 'trusted' | 'optimized' | 'adapted' — Trust maturity lifecycle stage for an agent.
    - `trusted_at` integer, nullable
    - `ttt_days` number, nullable
    - `dome_config_id` string, uuid, nullable
    - `dome_manually_configured` boolean
    - `import_metadata` union
      - ImportMetadata — Metadata about how an agent was imported.
        - `source` 'manual' — Import source for agent configuration.
        - `source_url` string, nullable
        - `imported_at` integer, nullable
        - `original_format` string, nullable
      - object
    - `capabilities` union
      - AgentCapabilities — Agent capability flags.
        - `streaming` boolean
        - `push_notifications` boolean
        - `code_execution` boolean
        - `memory` boolean
        - `delegation` boolean
        - `tools` string[]
      - object
    - `skills` union[]
      - union
        - AgentSkill — An agent skill definition (for A2A protocol agents).
          - `id` string, required
          - `name` string, required
          - `description` string, nullable
          - `input_modes` string[]
          - `output_modes` string[]
        - object
    - `latest_evaluation` AgentLatestEvaluation — Latest completed evaluation scores for an agent. Sideloaded on GET /agent-configurations when include_scores=true. Contains only domain-level score data — no UI or presentation concerns.
      - `evaluation_id` string, uuid, required
      - `overall_score` number, nullable
      - `reliability_score` number, nullable
      - `security_score` number, nullable
      - `safety_score` number, nullable
      - `completed_at` integer, nullable
  - `access_level` string, required
  - `warnings` string[]
  - `external_id` string, nullable
  - `identity_provider` string, nullable

## Other responses

- `422` — Validation Error

---

[API](https://skmtc.net/vijilai/apis/vijil-console-api-combined.md) · [All operations](https://skmtc.net/vijilai/apis/vijil-console-api-combined/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/vijilai/vijil-console-api-combined/versions/a8b2ee2b2d06/schema)