CtrlK
BlogDocsLog inGet started
Tessl Logo

spec-driven-development

Creates specs before coding. Use when starting a new project, feature, or significant change and no specification exists yet. Use when requirements are unclear, ambiguous, or only exist as a vague idea.

88

1.90x
Quality

Does it follow best practices?

Impact

95%

1.90x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

Evaluation results

97%

48%

Implementation Planning for New API Endpoints

Technical plan and task breakdown from approved spec

Criteria
Baseline
With context

Correct plan path

0%

100%

Correct todo path

0%

100%

Task has Description field

40%

70%

Task has Acceptance field

0%

100%

Task has Verify field

70%

100%

Task has Files field

70%

100%

Files count ≤5 per task

100%

100%

Dependency ordering

60%

100%

No implementation code

100%

100%

tasks/ directory created

0%

100%

90%

36%

Reliability Spec for the NewsSnap CLI

Spec from vague reliability requirement

Criteria
Baseline
With context

Assumptions block present

0%

100%

Multiple explicit assumptions listed

0%

100%

Vague requirement reframed

100%

100%

Measurable success criteria

100%

100%

No code written

100%

10%

Spec saved to file

100%

100%

Open Questions section

0%

100%

Objective section present

100%

100%

Boundaries section present

0%

100%

No unspecified features added

100%

87%

Commands section present

0%

100%

100%

51%

Specify a Dedupe Feature for a CSV Processing CLI

Spec document completeness for a new CLI feature

Criteria
Baseline
With context

Spec file saved

100%

100%

Objective section present

100%

100%

Commands section present

100%

100%

Full executable commands

100%

100%

Project Structure section present

0%

100%

Code Style section present

0%

100%

Code snippet in Code Style

0%

100%

Testing Strategy section present

0%

100%

Boundaries section present

0%

100%

Three-tier Boundaries

0%

100%

Success Criteria section present

100%

100%

Assumptions surfaced

70%

100%

No implementation code

0%

0%

Repository
addyosmani/agent-skills
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.