Automations
Automations are Intura's core capability: recurring AI agents that run content-strategy tasks on autopilot. You set one up once — what it should do, which knowledge it uses, and when it runs — then the agent works on schedule, researching, drafting, and delivering the result to you.
Runs on schedule
Daily, weekly, monthly, or manual.
Uses your knowledge
Grounded in your guidelines & audience.
Delivers results
To your inbox or a webhook.
Creating an Automation
Go to Workspace > Automations and click "New Automation" to start blank, or pick a template (e.g. "AI Search Answer Content (GEO)", "Daily TikTok Content Plan", "Weekly Content Performance Review") to pre-fill defaults. On the Setup tab, fill in three things:
Name
A clear title, e.g. "Daily Content Plan".
Brief
A one or two sentence description, in plain language, of the recurring task.
Instructions
The detailed prompt that defines the agent's role, mission, and the shape of the output. This matters most — write it in clear, decomposed steps.
Knowledge & Tools
Under Knowledge, pick the internal sources the agent may use: Content Guidelines, Target Audience, Knowledge Profile, your Instagram/TikTok accounts, Content Preferences (posts you've liked/disliked), Market Insight, or another automation's result. The agent fetches each source through a tool at run time — it is never pre-loaded into the prompt. Under Tools, you can enable verification tools: web search, fact-check, and a sentiment/cultural check.
Selections are enforced
Every knowledge category and verification tool you select becomes a required step: the agent must actually call it before finishing. This guarantees your guidelines and audience are genuinely used, not quietly ignored. Tip: enable Fact-Check for any claim-heavy output.
Schedule & Delivery
Schedule
Choose Daily (at a set time), Weekly (on a day), Monthly (on a date 1–28), or Manual only. The timezone is configurable — default GMT+7 (Indonesia). For manual automations, click "Run now" anytime to trigger one.
Delivery
Each time a run completes, the result can be sent by email (to an optional recipient list) or webhook (an HTTP POST to your endpoint, with optional HMAC signing). Without recipients, results still land in the Runs tab.
Runs & Results
Every execution is a "run". Open the Runs tab to see the history — each run has a status (queued → running → completed/failed), a markdown output, citations, a list of the knowledge actually used, a tool trace, and a delivery receipt. Click any run to open its detail. Each run spends 1 Carrot Point.
What a result looks like
A 'Weekly Content Performance Review' automation runs every Monday, pulls your tracked TikTok and Instagram analytics, web-searches for fresh angles, fact-checks its claims, and emails a tidy markdown report with citations to your team — while logging the full run in the Runs tab for reference.
Chaining Automations
One automation can build on another's latest result via the "Source Automation" knowledge category. For example, a "Content Generator" automation can ground itself on the daily output of a "Content Plan" automation — turning plans into finished content every day, automatically kept in sync.
Free plan: up to 2 automations
The Free plan lets you create up to 2 automations; paid plans remove the limit. Runs are async and have a strict time limit, so design each automation around one clear, well-scoped task rather than one giant catch-all job.
Was this page helpful?