Tools / Heuristic report

Heuristic report Pro

Turns findings against Nielsen's ten heuristics into a scored, exportable markdown report with duplicates merged across evaluators.

MCP tool name: build_heuristic_report. REST: POST /api/heuristic-report. Pro only.

Ask for it in plain words

Pro. Build a scored, exportable markdown report from heuristic evaluation findings you have written. Each finding has heuristic (1 to 10), severity (0 to 4), location, issue, effect, fix and optional effort (S, M, L) and evaluator. Merges duplicates across evaluators, computes a health score, and sorts a fix-first table.

Inputs

* required

Example

curl -X POST https://designers.allthepossibles.com/api/heuristic-report \
  -H "content-type: application/json" \
  -H "Authorization: Bearer dz_YOUR_KEY" \
  -d '{"title":"Cancel flow","task":"cancel a subscription","findings":[{"heuristic":9,"severity":3,"location":"Step 3 error","issue":"Something went wrong, no cause","effect":"people retry and then call support","fix":"name the cause and offer retry","effort":"S","evaluator":"a"},{"heuristic":9,"severity":4,"location":"Step 3 error","issue":"Something went wrong, no cause","effect":"x","fix":"y","evaluator":"b"},{"heuristic":3,"severity":2,"location":"Step 2","issue":"No way back","effect":"loses input","fix":"add a back link"}]}'

Example output

The report field, for the input above (structured fields come back alongside it):

# Cancel flow

Task evaluated: cancel a subscription
Evaluators: 2

**Health score: 82/100.** 2 distinct findings: 1 catastrophe, 0 major, 1 minor, 0 cosmetic, 0 not a problem. The score is 100 minus a penalty of 15, 8, 3 and 1 per finding at severity 4, 3, 2 and 1; it is a way to compare two rounds of the same evaluation, not a measure of absolute quality.

## Fix first

| # | Severity | Heuristic | Where | Issue | Effect | Fix | Effort |
|---|---|---|---|---|---|---|---|
| 1 | 4 catastrophe (2 evaluators) | 9. Help users recognize, diagnose, and recover from errors | Step 3 error | Something went wrong, no cause | people retry and then call support | name the cause and offer retry | S |
| 2 | 2 minor | 3. User control and freedom | Step 2 | No way back | loses input | add a back link | n/a |

## By heuristic

| Heuristic | Findings | Worst severity |
|---|---|---|
| 1. Visibility of system status | 0 | none |
| 2. Match between system and the real world | 0 | none |
| 3. User control and freedom | 1 | 2 minor |
| 4. Consistency and standards | 0 | none |
| 5. Error prevention | 0 | none |
| 6. Recognition rather than recall | 0 | none |
| 7. Flexibility and efficiency of use | 0 | none |
| 8. Aesthetic and minimalist design | 0 | none |
| 9. Help users recognize, diagnose, and recover from errors | 1 | 4 catastrophe |
| 10. Help and documentation | 0 | none |

## Method and limits

Findings were recorded against Nielsen's ten heuristics with severity 0 to 4 (0 not a problem, 1 cosmetic, 2 minor, 3 major, 4 catastrophe). A heuristic evaluation predicts problems; it does not replace watching people use the design. One evaluator finds roughly a third of problems and five find about three quarters, so treat a single-evaluator report as a first pass.

Limits of this tool

Deterministic: the same input always gives the same answer. It does not call a language model.