Objective Validator: checks an objective against its goal and key results
Objective Validator reviews one objective as part of your OKRs. A good objective says what you want, fits a bigger goal and can be measured through its key results. This agent checks all three and tells you what to fix. You run it on an objective, from a record action or as a workflow step, so the review is the same for every team.
What it does
- Reads the objective title and description and checks that they are clear and meaningful.
- If a goal is assigned, checks that the objective really supports that goal. If none is assigned and one clearly fits, it recommends the best available goal from the same division.
- Reads the key results linked to the objective and uses them to judge whether the objective can be measured.
- Marks the objective as valid when it needs no change.
- Otherwise lists only the fields that should change, each with a short explanation in plain labels.
When it helps
- A division head collects objectives from three teams. Running the agent shows which ones are not linked to any mid-term goal and which goal each one fits best.
- An objective sounds inspiring but has two key results that measure activity, not outcome. The validator points that out before the quarter starts.
- You add it as a check in an OKR workflow, so a new objective gets a first review before a human opens it.
How it works with your data
The agent has one tool, which reads records. It looks up the assigned goal, other goals of the same division and the key results of the objective, all within what the signed-in user is allowed to see. It does not write anything. Its answer is advice: a verdict and suggested field changes that you apply yourself. It uses a fast model, so a review comes back quickly.
Make it yours
This agent is a record in the Agents module. Change its instructions to reflect how your company writes objectives, for example a maximum length or a required link to a yearly goal. You can also change the model and temperature, duplicate it for a different team, or start from a blank agent. Worknano keeps the same UI and the same database across modules, so agents for OKRs, CRM or projects are edited in exactly the same place. Read Build your own agent and how to edit an agent.
Related: Key Result Validator checks each key result on its own. Chat Agent answers OKR questions in a thread. More background in AI agents.