CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

service-factory/gds-assessment-plugin

Granular 4-phase, 4-role GDS Service Standard assurance skills and synthesis engine.

Quality

87%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

past-assessments.mdrules/

Past assessments

Use the ingested GOV.UK service assessment reports as precedent for the assessment in front of you. Read a shortlist. Do not read the whole corpus. There are about 500 full reports, which is too large to load at once.

The catalogue is rebuilt by python scripts/build_past_assessment_catalogue.py from the plugin root. scripts/scrape_gds_reports.py rebuilds it after a scrape.

Before you score

  1. Read context/past-assessments/<phase>.json for the phase you are assessing (discovery, alpha, beta, or live).
  2. Choose at most 8 reports, then read those files under context/few-shot-examples/.
    • Prefer "standard": "14". Those use today's 14 points.
    • When the evidence looks thin on a point you own, prefer reports whose weak_points include that point (amber, red, or not-met).
    • When the evidence looks strong, include reports whose same points are green or met.
    • Include at least one report that failed a point you own and one that passed it, when both exist.
    • "standard": "18" is the old Digital Service Standard. You may read one for a phase narrative. Its point numbers are not today's points. Do not copy them into your score.
  3. The four curated exemplars (discovery-lead-stop.md, alpha-tech-spikes.md, beta-ur-fail-assistive-tech.md, beta-tech-pass.md) show how this plugin writes a verdict. They are not published panels. Do not spend the shortlist on them.
  4. If the catalogue file is missing, say so and judge from the evidence and the curated exemplar only.

Stop after 8 full reports even if more look relevant.

How to use what you read

Cite the published report when it bears on your decision: service title, phase, overall result, the point rating, and the source URL. Name the similarity in a sentence (same phase, same point, same kind of gap).

The evidence for this service decides the point. A published green rating does not clear a red line in context/rubrics/. In particular, red line B1 still marks Beta point 5 Not Met when no participant used assistive technology, whatever past reports did.

How published ratings translate

Published ratingIn this plugin
green, met, or met with adviceMet. Advice in a "met with advice" report is a recommendation, not a condition.
red, or not-metNot Met
amberNot Met when the report says evidence is missing and must be fixed before the next phase. Amber that is only optional advice is a recommendation.

Write your own decision as Met or Not Met. Do not write green, amber, or red in the report.

README.md

tile.json