CtrlK
BlogDocsLog inGet started
Tessl Logo

sharaf/migrate-to-tessl

Use when migrating, restructuring, publishing, or auditing an existing Claude skill into a Tessl tile; converting flat .md files or SKILL.md bundles; fixing Tessl Quality, Impact, Uplift, frontmatter, metadata, tile.json summary, README, markdown reference links, registry-vs-local Quality gaps, artifact anchors, auto-eval wait discipline, or pushing tile scores from 88-99% to 100%.

100

1.11x
Quality

100%

Does it follow best practices?

Impact

100%

1.11x

Average score across 4 eval scenarios

SecuritybySnyk

Advisory

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

task.mdevals/scenario-4/

Triage a Tessl Aggregate Score Regression

Background

A teammate migrated a release-note authoring skill into a Tessl tile. The team has already made several wording and README passes, and the latest version looks healthy in the usual summary commands. The registry total score is still below 100%, and the team is debating three possible explanations: a security advisory, missing manual verification, or another round of wording polish.

The team needs a concrete triage memo before making another publish. They also want the eval suite strengthened so this exact failure mode is caught in future migrations without making the scenario so obvious that a baseline agent gets it for free.

You have these inputs:

  • inputs/registry-search.json - registry search results for the target tile plus nearby comparison tiles
  • inputs/tile-info.txt - the current tile summary output
  • inputs/teammate-notes.md - hypotheses and local eval inventory

Use the Tessl migration skill guidance to diagnose the real blocker, reject the wrong fixes, and design a stronger eval scenario that would catch this issue without leaking the answer in the task prompt.

Output Specification

Write two files:

  1. score-triage.md

    • root cause
    • evidence from the provided artifacts
    • wrong hypotheses to avoid
    • minimum fix plan
    • verification commands and fields
  2. scenario-repair-plan.md

    • why the current failure mode deserves its own eval scenario
    • how to make that scenario harder than a direct "why is this 93%" prompt
    • proposed fixture files
    • proposed task framing
    • proposed weighted rubric checks totaling 100 points

evals

README.md

tile.json