Score skill execution quality with a 5-dimension rubric and update the local registry.
npx skills add https://github.com/Shmayro/singularity-claude --skill scoring-shmayro
Please help me install this Agent Skill. Skill: scoring Source: https://github.com/Shmayro/singularity-claude/tree/main/skills/scoring Command: npx skills add https://github.com/Shmayro/singularity-claude --skill scoring-shmayro
This skill provides a structured score and telemetry after each skill execution, enabling continuous quality improvement of singularity-managed skills.
After a skill completes, run the scoring workflow to evaluate its performance and update the score registry.
High-intent search queries and answers about installing and using this skill.
To evaluate skill performance, run the scoring workflow to apply a 5-dimension rubric. This generates a 0-100 score to rate execution quality and compare results across multiple tracked runs.
A 5-dimension scoring rubric for execution assessment rates performance 0-20 per dimension across Correctness, Completeness, Edge Case Handling, Efficiency, and Reusability. These dimensions combine to produce a comprehensive 0-100 score.
Trigger repair or crystallization by using threshold-driven maintenance rules based on assessment scores. The scoring system monitors score averages and run counts to automatically initiate these maintenance decisions.
To record evaluation results and emit telemetry locally, the scoring skill writes results to the local registry and emits telemetry through provided scripts. This enables continuous quality improvement and auditing.
The scoring workflow requires no external dependencies to assess execution quality. It operates entirely through local scripts and references to evaluate tracked skills and update the registry.
You need continuous quality monitoring for tracked skills to enable structured scoring and telemetry after each execution. This process identifies performance degradation over time and triggers necessary maintenance actions.
ShmayroDiagnose and rewrite underperforming singularity-managed skills with iterative scoring validation.
ShmayroAnalyze score history, registry data, and telemetry to report skill health status.
PixelSculptorGenerate project structures from tech stack definitions and log verification.
tnezEvaluate agent skills across Clarity, Completeness, Examples, and Focus using a rubric-based assessment.
