TokPortal + OpenClaw

TokPortal has an OpenClaw integration path through ClawHub. The important technical detail is that the skill exposes the same public TokPortal API surface as the MCP server: bundles, account configuration, videos, webhooks, uploads, accounts, and analytics.

Install

clawhub install tokportal

Skill page: clawhub.ai/naybu256/tokportal

Configure

Set a TokPortal API key from the Developer Portal.

Depending on your OpenClaw setup, configure the key either in the skill config or as an environment variable:

TOKPORTAL_API_KEY=sk_your_key_here

What the agent can do

The agent can only do what the public API supports. It can call generated tools such as:

Tool	API surface
`tokportal_get_credit_balance`	`GET /credits/balance`
`tokportal_create_bundle`	`POST /bundles`
`tokportal_create_bundles_bulk`	`POST /bundles/bulk`
`tokportal_configure_bundle_account`	`PUT /bundles/{id}/account`
`tokportal_configure_bundle_video`	`PUT /bundles/{id}/videos/{position}`
`tokportal_batch_configure_bundle_videos`	`PUT /bundles/{id}/videos/batch`
`tokportal_publish_bundle`	`POST /bundles/{id}/publish`
`tokportal_get_account_analytics`	`GET /accounts/{id}/analytics`

See the full MCP tool list.

Safe prompt pattern

Ask the agent to plan, show the API payload, then wait for confirmation before calling mutating tools:

"Create a draft plan for a US TikTok bundle with 5 videos and Advanced Niche Warming on 6 targets. Show the exact tokportal_create_bundle payload and estimated credit cost. Do not execute until I confirm."

Then approve once the payload matches the API docs:

{
  "bundle_type": "account_and_videos",
  "platform": "tiktok",
  "country": "US",
  "videos_quantity": 5,
  "wants_advanced_warming": true,
  "advanced_warming_terms": [
    "gym motivation",
    "home workout",
    "protein recipes",
    "beginner lifting",
    "hiit routine",
    "fitness transformation"
  ],
  "external_ref": "openclaw-fitness-us"
}

wants_advanced_warming: true is mandatory whenever the payload carries advanced_warming_terms or advanced_warming_terms_count; TokPortal never infers it and rejects the call with ADVANCED_WARMING_FLAG_REQUIRED. Targets are 3-30, in multiples of 3, each 2-50 characters. At the standard rate of 5 credits per target, the 6 targets above add 30 credits on top of the 32 for the account and 2 per video slot, so this bundle is 32 + (5 x 2) + 30 = 72 credits. Have the agent confirm the figure with tokportal_get_credit_costs instead of quoting it from memory.

Realistic workflow

Check credits with tokportal_get_credit_balance.
Create a bundle with tokportal_create_bundle.
Configure account profile fields with tokportal_configure_bundle_account.
Configure videos from public URLs or uploaded media with tokportal_configure_bundle_video.
Publish with tokportal_publish_bundle.
Monitor bundle status, webhook events, or analytics.

The agent should not promise organic reach, view counts, or platform enforcement outcomes. TokPortal exposes operational workflow tools; performance still depends on content, platform behavior, timing, and account state.