CtrlK
BlogDocsLog inGet started
Tessl Logo

shape

Resolves technical HOW decisions — architecture choices, technology selection, and design patterns. Use shape when the intent is clear but the technical path is not. Triggers on "shape this", "architecture for X", "how should I build", "which pattern", "technical approach".

94

1.72x
Quality

Does it follow best practices?

Impact

83%

1.72x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

Evaluation results

93%

7%

Shape the Loop Architecture for a Contract Review Pipeline

Loop architecture handoff card

Criteria
Without context
With context

Verifier locus addressed

100%

100%

Recommends best-of-N or fixed-round

100%

100%

Rejects self-improving loop

100%

100%

Judgment-delegated termination rejected

100%

80%

Hard cap or fixed termination rule

100%

100%

Plan mutability addressed

62%

62%

Checksum first

12%

100%

Carry-forward last

50%

100%

No temporal language

100%

75%

Paths ruled out included

100%

100%

No empty scaffolding

100%

100%

81%

81%

Shape the Developer Scaffolding CLI

Decision prompt formatting and batching

Criteria
Without context
With context

Batched independent questions

0%

100%

Exactly 3 candidate options

0%

100%

Gather facts escape hatch present

0%

100%

No markdown in previews

0%

70%

ALL CAPS section headers in previews

0%

80%

Short preview lines

0%

50%

Label is recommendation first

0%

37%

No technique names in output

0%

100%

Self-contained question UI

0%

66%

Brief scannable output

0%

75%

76%

16%

Caching Strategy: Shape and Handoff

Handoff card quality: gate audit, stranger test, no volatile refs

Criteria
Without context
With context

Checksum first

0%

0%

No volatile references

0%

90%

No empty sections

100%

90%

No provenance markup

100%

100%

Timeless phrasing

87%

62%

Stranger test

90%

100%

Carry-forward section last

25%

62%

Gate verdict first

0%

50%

Deletion pass evidence

100%

100%

Completeness pass evidence

100%

100%

Repository
saadshahd/moo.md
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.