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
100%
Does it follow best practices?
Impact
100%
1.11xAverage score across 4 eval scenarios
Advisory
Suggest reviewing before use
Evaluation results
100%
11%Maximize Tile Quality Before Registry Submission
Tile quality audit and frontmatter fix
Non-standard keys removed
100%
100%
Metadata block added
100%
100%
String-only metadata values
100%
100%
metadata.version key
100%
100%
Top-level version removed
100%
100%
Use when clause
26%
100%
tile.json summary updated
100%
100%
README.md created
100%
100%
README includes when-to-use
100%
100%
Audit report created
100%
100%
100%
Write a Tessl Migration Runbook for the Platform Team
Migration runbook and preflight guide
CLI version check
100%
100%
Login check
100%
100%
Required skills check
100%
100%
Project creation timing
100%
100%
Eval caching pitfall
100%
100%
Activation vs content eval
100%
100%
Default visibility private
100%
100%
Default version 0.1.0
100%
100%
Default eval agent
100%
100%
Dry-run before publish
100%
100%
Minor bump for final publish
100%
100%
Stop conditions listed
100%
100%
100%
23%Prepare a Skill Bundle for Tessl Publication
Tessl skill bundle restructuring
Markdown links for refs
100%
100%
No backtick paths remain
100%
100%
Standard top-level frontmatter
100%
100%
Metadata block present
0%
100%
Metadata string values only
0%
100%
metadata.version present
0%
100%
Use when clause in description
100%
100%
tile.json summary synced
100%
100%
README at tile root
100%
100%
Reference files in correct location
100%
100%
100%
5%Triage a Tessl Aggregate Score Regression
Aggregate score triage and scenario repair
Writes requested files
100%
100%
Identifies eval count as blocker
100%
100%
Cites exact count and aggregate
100%
100%
Recognizes perfect score-bearing dimensions
100%
100%
Explains three-scenario threshold
100%
100%
Uses comparison evidence
100%
100%
Rejects wrong hypotheses
100%
100%
Prescribes adding a real third scenario
100%
100%
Includes publish and auto-eval wait
50%
100%
Verifies with registry search JSON
100%
100%
Uses tile info appropriately
100%
100%
Identifies answer leakage risk
100%
100%
Designs harder fixtures
100%
100%
Proposes concrete task framing
100%
100%
Proposes weighted rubric totaling 100
80%
100%
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6