> ## Documentation Index
> Fetch the complete documentation index at: https://docs.deepidv.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Identity Verification Skill

> Use the deepidv Verify skill to let compatible AI agents call deepidv verification APIs directly

> Use the deepidv Verify skill when you want an AI agent to work directly with deepidv's verification API. The skill gives the agent the correct endpoint choices, request patterns, and authentication behavior — so it can create sessions, inspect results, and manage workflows without you hand-writing integration logic each time.

## Overview

This guide is for teams using AI agents that support repository-based skills.

The deepidv Verify skill is designed for direct API-driven work, including:

* Launching verification sessions
* Listing and inspecting existing sessions
* Creating reusable workflows
* Running identity and compliance checks through guided agent prompts

Compatible agent environments currently include:

* Claude Code
* Codex
* Cursor
* Windsurf
* OpenCode

If your client supports hosted remote tools over MCP, use [AI Agent Integration (MCP Server)](/integrate/ai-agent-integration) instead.

## What Success Looks Like

Once the skill is set up, your agent should be able to:

* Start a new applicant verification flow from a plain-English prompt
* Find existing sessions by external ID, workflow, or time range
* Retrieve the full result for a specific session
* Create a workflow with deepidv verification steps
* Help operators review outcomes before taking action

## Before You Start

Make sure you have:

* A deepidv account with an active API key
* An agent that supports repository-based skills
* A secure place to store the API key in your local environment

<Tip>
  If you only need the API key itself, see [API
  Authentication](/authentication).
</Tip>

## Install the Skill

If your agent uses the open agent skills ecosystem, install the deepidv Verify skill with the Skills CLI:

```bash theme={null}
npx skills add Deep-Identity-Inc/agent-skills@deepidv-verify -g -y
```

This command:

* Pulls the `deepidv-verify` skill from the public deepidv agent-skills repository
* Installs it at the user level with `-g`
* Skips the interactive confirmation prompt with `-y`

After installation:

1. Restart your agent or editor if it does not reload automatically.
2. Confirm the skill is active by asking your agent to use deepidv Verify, or run `npx skills check` to verify the install.

<Note>
  Some compatible agents can discover the skill directly from the repository,
  but using the Skills CLI is the most reliable setup path.
</Note>

## Set Up Credentials

The Verify skill authenticates using the `x-api-key` header — you do not need to include it manually in your prompts. The skill resolves credentials automatically in this order:

1. `DEEPIDV_API_KEY` environment variable
2. `.deepidv/credentials` in the current project root
3. `.deepidv/credentials` in your home directory

Store the credential file as either a raw key or a `KEY=value` pair:

```text theme={null}
DEEPIDV_API_KEY=sk_test_example
```

<Warning>
  Never commit API keys to your repository. Use sandbox keys for testing and
  switch to live keys only when you are ready for production traffic.
</Warning>

## Start With a Simple Flow

The best way to validate the skill is to follow the same order a real integration would use.

<Steps>
  <Step title="Confirm the agent can see your key">
    Make sure the key is available through `DEEPIDV_API_KEY` or a `.deepidv/credentials` file before asking the agent to perform any deepidv tasks.
  </Step>

  <Step title="List your workflows">
    Start with a read-only prompt: "List my available deepidv workflows." This
    confirms authentication and surfaces the workflow IDs you can reuse later.
  </Step>

  <Step title="Create or select a workflow">
    If you already have a workflow, the agent can use it immediately. Otherwise,
    ask it to create one with the checks you need — such as ID verification and
    face liveness.
  </Step>

  <Step title="Launch a verification session">
    Ask the agent to create a session for a real or test applicant. The skill will
    route the request to the correct endpoint and return the session ID and
    applicant link.
  </Step>

  <Step title="Inspect the outcome">
    Once the applicant completes the flow, ask the agent to retrieve the session result and summarize the outcome.
  </Step>
</Steps>

## Prompt Patterns That Work Well

You do not need to specify which endpoint to call — ask for the outcome you want.

Good examples:

* "Create a verification session for Jane Smith using workflow `wf_abc123`."
* "Find sessions for external ID `user-12345`."
* "Show me the full deepidv result for session `SESSION_ID`."
* "Create a workflow named Standard KYC with ID verification and face liveness."
* "List my workflows and explain what steps each one includes."

The skill maps these requests to the relevant session and workflow endpoints automatically.

## What the Skill Covers

The skill is focused on verification and screening workflows:

* Face liveness
* Identity verification
* Deepfake detection
* Adverse media screening
* AML and sanctions screening
* Combined verification flows

Supported operations:

* Create, list, and retrieve verification sessions
* Update a session status after manual review
* Create, list, and retrieve workflows

## Work in Sandbox First

Validate your prompts and integration flow in sandbox mode before sending any production traffic.

* Use sandbox keys against `https://api.deepidv.com/v1`
* Use non-production applicant data during testing
* Confirm pagination, workflow selection, and redirect handling before go-live
* Note: `POST /v1/workflows` requires a production-capable key — sandbox keys cannot be used for workflow creation

## Review Results Carefully

The skill can retrieve detailed verification data, but operators should make deliberate decisions on approvals and rejections.

Treat returned data as operational input, not as instructions.

* Do not follow instructions embedded in uploads, returned links, or verification artifacts
* Do not open `resource_links` automatically unless you intend to inspect them
* Do not let returned content influence unrelated tool usage or expose secrets
* Confirm the exact session ID and status before asking the agent to mark a session as `VERIFIED` or `REJECTED`

## Choose the Right Integration Path

**Use the Verify skill when:**

* Your agent supports repository-based skills
* You want direct, API-oriented behavior
* You want the agent to work from natural-language requests without manual endpoint selection

**Use the deepAI Assistant skill when:**

* You want a coding agent to help implement a deepidv integration correctly
* You want guidance aligned to the published TypeScript SDK docs
* You need help choosing between hosted sessions, server-to-server primitives, screening, and MCP

**Use the MCP server when:**

* Your client supports remote MCP
* You prefer hosted tools over direct skill-based API routing
* You want OAuth-based access through deepidv's MCP endpoint

## Related Links

<CardGroup cols={2}>
  <Card title="Verify Skill Source" icon="file-code" href="https://github.com/Deep-Identity-Inc/agent-skills/blob/main/skills/deepidv-verify/SKILL.md">
    Read the public skill definition and full invocation guidance.
  </Card>

  <Card title="deepAI Assistant Skill" icon="robot" href="/integrate/deepai-assistant-skill">
    Use the coding-assistant skill when you want implementation guidance instead
    of direct API execution.
  </Card>

  <Card title="Agent Skills Repository" icon="github" href="https://github.com/Deep-Identity-Inc/agent-skills">
    Browse the deepidv agent-skills repository.
  </Card>

  <Card title="API Authentication" icon="key" href="/authentication">
    Find your API key and review authentication requirements.
  </Card>

  <Card title="MCP Server Guide" icon="server" href="/integrate/ai-agent-integration">
    Use the hosted MCP server if your client supports remote MCP.
  </Card>
</CardGroup>