Agent readiness assessment
Is This Task Ready for an Agent?
Check whether one recurring business task has the practical foundations for a focused, reviewable AI-assisted workflow.
- Audience
- Business, technology, governance, operations, and PMO leaders
- Outcome
- A practical next step for preparing or designing one task
Assess one task, not the whole organization
Choose the answer that best reflects the work today. The questions apply to PMO work such as a business-case review or portfolio briefing, and to governance, operations, procurement, finance, and technology work with a repeatable outcome.
This assessment is a planning tool, not a certification or a guarantee that a task is suitable for automation.
It runs only in your browser. Do not enter task names, documents, people, or other sensitive information; answers are not stored, transmitted, or added to the URL.
What the assessment checks
A focused agent is most useful when the work repeats, the job is clear, sources are known, and a person owns the result. It also needs an honest uncertainty path: when evidence is absent or conflicting, the workflow should say so, ask for clarification, or escalate instead of guessing.
The score is transparent: established earns 2 points, partly defined earns 1 point, and not established earns 0 points. A high score is a signal to design a bounded first use case—not permission to skip source validation, testing, or human review.
How to improve an early-stage task
- Make the recurring job visible. Name the trigger, intended user, expected output, and decision or action it supports.
- Set the source boundary. Identify permitted records, the owner responsible for them, and what the workflow must ignore.
- Keep the human operating model explicit. Name the reviewer, approval or action owner, exception path, and success measure.
- Test against real work. Use incomplete and conflicting examples, inspect where a human changes the output, and improve the design before expanding it.
Why readiness is not automation approval
A task can be ready for a focused design while still requiring strict human review. The assessment does not decide whether AI may access information, make a decision, or take an action. Those boundaries depend on the work, the evidence, the risk, and accountable people.