Run a Method and Judge the Result Properly

Completion is not enough

The system should treat quality as the real bar, not just completion status. A method can finish and still produce output that is too generic to trust.

• A method is not done if the output is weak or generic.
• A passing run should still be reviewed for grounding and usefulness.
• The real question is whether the report retrieved the information it was meant to retrieve.

What to look for

Use the output itself to judge whether the method performed correctly.

• Does the report reflect the project context accurately?
• Does it use evidence rather than placeholder logic?
• Does it produce method-specific reasoning rather than a generic shell?

When to rerun

If context changes materially, the method may need to be rerun so the evidence base stays current.

• Big discovery changes can make earlier reports stale.
• Methods should improve when richer project intelligence becomes available.

Use this page to review report quality

Before you move into the next decision step, make sure the method output is specific, evidence-backed, and worth carrying into the wider Flowlytics workflow.