AI in Workflows

KnoxCall brings AI into workflows two ways: an AI step you drop into a graph like any other node, and an AI copilot that drafts workflows from a plain-language prompt and explains failed runs.

The AI Step

The AI node runs an Anthropic model with structured output: each operation declares the shape it returns, and you get typed data back with no prompt-parsing glue. Five operations:

Operation	What it does	Key config	Output
Classify	Picks one label from a list	`categories: [...]`	`category`, `confidence`, `reasoning`
Extract	Pulls named fields out of text	`fields: [{ name, description, type? }]`	`fields` (an object)
Summarize	Condenses text to about a target length	`maxWords` (default 80)	`summary`
Generate	Free-form writing	—	`text`
Translate	Translates into another language	`targetLanguage` (e.g. `"French"`, `"pt-BR"`)	`text`, `sourceLanguage`

Every operation takes an input (the text to work on, or the prompt for Generate) and an optional instruction for extra guidance. Both support {{ ... }} variables, and so does targetLanguage. Example — classify an inbound message:

{
  "type": "ai",
  "data": {
    "type": "ai",
    "config": {
      "operation": "classify",
      "input": "{{trigger.data.message}}",
      "categories": ["billing", "technical", "sales"]
    }
  }
}

Reference the result downstream by the node’s id, for example {{ai_1.category}}. Example — translate a reply into the customer’s language:

{
  "operation": "translate",
  "input": "{{draft_1.text}}",
  "targetLanguage": "{{trigger.data.customer.locale}}"
}

Output length

Max Output Tokens caps how much the model may write, on every operation. The default is 1,024 and the maximum is 16,000. The cap bounds what a single step can spend and how long it holds the run, so raise it only for steps that genuinely produce long output, such as a long extraction or a full-length draft. A value set through a {{variable}} is checked when the run happens: anything that isn’t a number falls back to 1,024, and anything out of range is clamped.

Model & Credentials

AI steps run on your own Anthropic key. Add it under Settings → Integrations → Anthropic. There is no fallback to a KnoxCall key: without yours, the step fails with a message telling you to add one. The model is configured centrally, not per node; the default is claude-sonnet-5. Calls go through the KnoxCall AI Gateway, so an AI step gets the same budgets, firewall and usage attribution as the rest of your AI traffic.

Metering & Limits

AI nodes count as weighted operations (×3) against your monthly operations quota.
AI-token usage is metered against a monthly AI-token budget per plan tier.
When either limit is reached, an AI step is blocked before it spends anything, rather than overspending.

The AI Copilot

Open the AI panel in the builder toolbar to draft a workflow from a description.

Draft a workflow from a prompt

Describe what you want — for example:

“When a GitHub issue is opened, classify it as bug or feature, and post a Slack message to the right channel.”

The copilot returns a coherent, auto-laid-out graph (trigger → nodes → edges) that lands unsaved on your canvas. It goes through the same publish gate as a hand-built workflow: structural validation is enforced, and any per-node configuration gaps come back as warnings so you know exactly what to fill in (connections, secrets, endpoints) before publishing.

Explain a run

On any finished execution, click Explain in the run log. The copilot reads the run’s redacted step history and returns a plain-language explanation of what happened plus a suggested fix — for example, diagnosing an unreachable URL and pointing at the exact node to correct.

Metering

Copilot calls are rate-limited and draw from the same AI-token budget as AI steps (each call reports its own token cost).

Tips

Start Extract/Classify with a few clear labels or fields; add more only if results are ambiguous.
Use the copilot to scaffold, then refine node config by hand — it’s a starting point, not a black box.
Keep an Anthropic key in Integrations if you want AI usage billed to your own account.

​AI in Workflows

​The AI Step

​Output length

​Model & Credentials

​Metering & Limits

​The AI Copilot

​Draft a workflow from a prompt

​Explain a run

​Metering

​Tips

​Next Steps

Creating Workflows

Templates