rubrkit
Pricing
Sign In

The instrument

Turn loose instructions into bounded, testable procedures.

The rubr_flow DSL plus a validator that catches vague conditions, missing stop rules, and absent output contracts.

Validate a rubr_flow file

rubr_flow blocks

TASK
CONTEXT
INPUTS
RULES
FLOW
OUTPUT
VERIFY
rubr_flow

A native format for instructions an agent can actually follow.

For teams that need stricter control, rubr_flow turns loose intent into a compact procedure an agent can follow.

Explicit work order
Bounded edits
Pass/fail verification
Before

Intent without machinery

Review our onboarding flow and fix anything confusing.

After

rubr_flow procedure

TASK "Improve onboarding completion"
CONTEXT
  user is new to [PRODUCT]
  primary action is [TARGET ACTION]
INPUTS
  current_flow = app screens
  analytics_notes = drop-off data
  support_themes = user confusion reports
RULES
  change only copy and step order
  preserve required legal text
ON missing_context
  ASK user "Which onboarding detail is missing?" -> missing_detail
FLOW
  REVIEW each screen -> friction_notes
  RANK issues by user impact
  EDIT the highest-impact issue
OUTPUT
  changed_copy: final text
  rationale: why this improves completion
  risk_notes: constraints preserved
VERIFY
  PASS WHEN user can identify the next action in one pass

This is not another prompt type. It is the control surface for turning intent into bounded agent work.

Run an audit
The seven blocks

What makes a procedure testable.

rubr_flow is pseudocode for instructions. Each block forces a decision that loose prose leaves implicit — which is exactly what lets you write a check against the result.

TASK · CONTEXT · INPUTS

What the agent is doing, what it may assume, and exactly what it receives.

RULES · FLOW

The boundaries it must respect and the ordered steps it follows — with stop rules, so it cannot wander.

OUTPUT · VERIFY

The shape of a correct result and the check that says whether the result is actually correct.

Validated before it runs

The validator flags vague conditions, missing stop rules, and absent output contracts before an agent ever follows the procedure.

FAQ

Answers before you start.

Explicit inputs, bounded steps, a stop rule, and an output contract — so you can write a check that says whether the result is correct. Loose prose has none of these, which is why it drifts. rubr_flow makes them mandatory.

Know which instructions are ready to run.

Validate a rubr_flow file
Newsletter

Follow the review loop as it ships.

Notes on AI artifact testing, rubr_flow conversion, evals, and proof reports.