Krafthaus workflow app

AI Agent Action Gate

A concrete AI agent action boundary for software companies: intake the proposed action facts, run a direct_declarative_rulebook verdict on Decide, store the Decision Record, and hand the execution owner the next action before autonomous software acts.

Generate agent proposal Run demo Open manifest Read app contract

Buyer wedge

Before autonomous software sends, changes, refunds, or updates.

This is the direct version of the broader Krafthaus thesis: before a customer-facing agent action executes, Krafthaus captures the facts, Decide returns a binding verdict, and the owner gets an action packet. It is not asking an LLM for a loose business judgment.

Sales

AI sales agent send boundary

Gate a customer-facing agent action before pricing, discount language, or next-step commitments leave the company.

Support

AI support action boundary

Route refunds, subscription changes, and sensitive customer updates when confidence, consent, or approval is missing.

Ops

AI back-office action boundary

Allow low-risk reversible updates, but require review for financial, security, or irreversible actions.

01 Action packet

A running ai_agent_action_gate_packet with intake facts, verdict material, handoff, packet hash, and Decision Record references.

02 Rulebook-backed action logic

Direct declarative rules for confidence, consent, human approval, value at risk, reversibility, and customer impact.

03 Execution handoff

Execute the agent action, route it to human review, or block execution until the risk clears.

Executable demo

Run an agent action through the gate.

This form posts closed action facts to /api/ai-agent-action-gate. The server validates the intake, calls Decide with a pinned direct_declarative_rulebook, and only returns an ai_agent_action_gate_packet when the Decision Record material is present.

Action ID Agent name Action type Customer impact Monetary value USD Policy confidence Sensitive data access User consent present Human approval present Reversible Execution owner

Runtime: direct_declarative_rulebook

Decision Record material

AI Agent Action Packet

Ready to run

Submit an action to see application_verdict, action, matched rule, runtime binding, rulebook_attestation material, and the returned ai_agent_action_gate_packet with packet_hash.

Shareable packet URL

Application contract

A registered Krafthaus app surface for AI action control.

AI Agent Action Gate is a software-company operating workflow that shows how Krafthaus can install Decide between a proposed autonomous action and the system that would mutate customer, revenue, or operational state.

Manifest ai-agent-action-gate-v1.json

Application Manifest v1 pins the workflow, intake contract, direct rulebook binding, record fields, action boundary, and handoff.

Rulebook rulebook-v1.schema.json

The production core evaluates limited declarative JSON. Customer executable rulebooks are rejected.

Sprint Start from AI agent action

Use the proposal template to tailor this boundary to a company's AI agent actions, approval policy, and handoff owners.