HNK SIGNALLESS NOISE. BETTER DECISIONS. Search

THE HNK STANDARD

Show the work.
Explain the decision.

A useful recommendation makes its evidence and limitations easy to find.

What we require from a test

  1. Define the use case, audience and success criteria.
  2. Record the version, plan, date, settings and representative inputs.
  3. Keep the output and a baseline; record time spent and revision effort.
  4. Evaluate quality, costs, ease of use and failures against the task.
  5. Publish evidence, limitations and the reasoning behind the verdict.

Repeated runs matter for reliability. A single successful demo does not establish consistent results. Results apply to the conditions tested.

Every badge has a meaning.

TESTED

A real HNK test is complete and its evidence is published.

TESTING

A real HNK test has started; results are not final.

RECOMMENDED

Curated by HNK, without a completed full HNK test.

LISTED

A directory entry. No recommendation or test result is implied.

The HNK SIGNAL score

Only TESTED tools can receive a score from 0 to 10. Every scored review must show the measurements and reasoning. There are no scores in the current directory.

01

Output quality

Does the output meet the task’s stated criteria?

02

Time saved

Time to a usable result, including setup and revisions, against a baseline.

03

Ease of use

Friction, learning effort and manual intervention.

04

Value

Total task cost, restrictions and practical usefulness.

05

Reliability

Consistency across repeated runs and how failures are handled.

Overall weighting and any verdict — WORTH IT, SKIP or DEPENDS — must be explained in the review.

Commercial relationships

Affiliate availability does not determine a score. Sponsored content and active affiliate links must be disclosed alongside the content.

Read the disclosure →