Define process measures and controls with AI
Create a KPI register and a separate control register from a validated process narrative. Keep each measure’s source and cadence distinct from evidence that a control operated.
Define process measures and controls with AI when a team has mapped the work but cannot yet review it consistently. The okhp3-process-measures-controls skill helps document what to measure, who owns it and how frequently it is collected. Controls have their own purpose, evidence and response when a check fails.
For a procurement process, purchase approval time is a measure. An approval requirement before payment is a control. The first needs a calculation and source. The second needs a responsible role and evidence that the check took place. Unknown numeric targets should stay open until the team agrees them.
What it does
- Define cycle time, error rate or cost per transaction.
- Tie a proposed KPI to its operational data source.
- Separate preventive and detective controls.
- Identify the evidence expected from each check.
- Escalate missing targets or ownership for confirmation.
Install and use
- Install the skill and provide a validated structured process narrative and any agreed measures. Follow the documented manual workflow or use its exported helper with parsed inputs. The local helper uses Node.js.
- Ask, “Create measures and controls registers for purchase approval. Keep missing targets open and show the evidence for each control.”
- Review both registers with their owners. The command-line wrapper currently reports counts from empty input lists, so verify the populated outputs against the narrative.
FAQ
Source
Made by OverKill Hill P³. Original skill: okhp3-process-measures-controls, published on GitHub under the MIT licence.
View on GitHub