Skip to content

feat: improve skill scores across 12 skills#164

Open
popey wants to merge 1 commit intoastronomer:mainfrom
popey:improve/skill-review-optimization
Open

feat: improve skill scores across 12 skills#164
popey wants to merge 1 commit intoastronomer:mainfrom
popey:improve/skill-review-optimization

Conversation

@popey
Copy link

@popey popey commented Mar 16, 2026

Hullo @schnie 👋

I ran your skills through tessl skill review at work and found some targeted improvements. These are the ten most improved:

score_card

Here's the full before/after:

Skill Before After Change
deploying-airflow 77% 100% +23%
profiling-tables 77% 94% +17%
managing-astro-deployments 85% 100% +15%
setting-up-astro-project 89% 100% +11%
warehouse-init 81% 89% +8%
airflow-adapter 81% 89% +8%
annotating-task-lineage 86% 93% +7%
creating-openlineage-extractors 86% 93% +7%
tracing-downstream-lineage 88% 93% +5%
troubleshooting-astro-deployments 88% 93% +5%
testing-dags 84% 89% +5%

Skills already at 90%+ were left untouched — they're excellent as-is.

Changes made

Description improvements (all changed skills)

  • Expanded frontmatter descriptions with specific concrete actions (e.g., "Configure deployment pipelines, manage DAG-only and full image deploys")
  • Added natural trigger term variations users would actually say (e.g., "astro deploy", "catalog tables", "scan database schema")

deploying-airflow (+23%)

  • Consolidated verbose deploy command explanations into compact, scannable format
  • Added verification steps after each deployment method (astro deployment inspect, curl health, kubectl get pods)
  • Trimmed explanatory prose Claude doesn't need

profiling-tables (+17%)

  • Added large table warning with TABLESAMPLE guidance upfront
  • Consolidated data quality dimensions from prose into a compact reference table
  • Trimmed "This reveals:" explanatory bullets

managing-astro-deployments (+15%)

  • Removed intro prose and "What inspect Shows" section (discoverable via --help)
  • Condensed auth and workspace sections into compact command blocks

setting-up-astro-project (+11%)

  • Added explicit validation steps (astro dev run airflow connections list, variables list)
  • Removed intro prose, added verify-after-restart guidance

warehouse-init (+8%)

  • Removed redundant "Scripts are relative to" comments after first mention
  • Condensed "What This Does" section into a single-line summary

airflow-adapter (+8%)

  • Added concrete 3-step workflow for adding new API methods with code examples

annotating-task-lineage & creating-openlineage-extractors (+7% each)

  • Added verification sections with steps to confirm lineage appears correctly
  • Condensed "On Astro" marketing into a single-line note

testing-dags, tracing-downstream-lineage, troubleshooting-astro-deployments (+5% each)

  • Tightened verbose instruction blocks
  • Removed intro prose that restated the skill title

Honest disclosure — I work at @tesslio where we build tooling around skills like these. Not a pitch - just saw room for improvement and wanted to contribute.

Want to self-improve your skills? Just point your agent (Claude Code, Codex, etc.) at this Tessl guide and ask it to optimize your skill. Ping me - @popey - if you hit any snags.

Thanks in advance 🙏

Hullo @schnie 👋 I ran your skills through `tessl skill review` at work and found some targeted improvements. Here's the full before/after: | Skill | Before | After | Change | |-------|--------|-------|--------| | deploying-airflow | 77% | 100% | +23% | | profiling-tables | 77% | 94% | +17% | | managing-astro-deployments | 85% | 100% | +15% | | setting-up-astro-project | 89% | 100% | +11% | | warehouse-init | 81% | 89% | +8% | | airflow-adapter | 81% | 89% | +8% | | annotating-task-lineage | 86% | 93% | +7% | | creating-openlineage-extractors | 86% | 93% | +7% | | tracing-downstream-lineage | 88% | 93% | +5% | | troubleshooting-astro-deployments | 88% | 93% | +5% | | testing-dags | 84% | 89% | +5% | Skills already at 90%+ were left untouched — they're excellent as-is. <details> <summary>Changes made</summary> **Description improvements (all changed skills)** - Expanded frontmatter descriptions with specific concrete actions (e.g., "Configure deployment pipelines, manage DAG-only and full image deploys") - Added natural trigger term variations users would actually say (e.g., "astro deploy", "catalog tables", "scan database schema") **deploying-airflow (+23%)** - Consolidated verbose deploy command explanations into compact, scannable format - Added verification steps after each deployment method (`astro deployment inspect`, `curl health`, `kubectl get pods`) - Trimmed explanatory prose Claude doesn't need **profiling-tables (+17%)** - Added large table warning with TABLESAMPLE guidance upfront - Consolidated data quality dimensions from prose into a compact reference table - Trimmed "This reveals:" explanatory bullets **managing-astro-deployments (+15%)** - Removed intro prose and "What inspect Shows" section (discoverable via `--help`) - Condensed auth and workspace sections into compact command blocks **setting-up-astro-project (+11%)** - Added explicit validation steps (`astro dev run airflow connections list`, `variables list`) - Removed intro prose, added verify-after-restart guidance **warehouse-init (+8%)** - Removed redundant "Scripts are relative to" comments after first mention - Condensed "What This Does" section into a single-line summary **airflow-adapter (+8%)** - Added concrete 3-step workflow for adding new API methods with code examples **annotating-task-lineage & creating-openlineage-extractors (+7% each)** - Added verification sections with steps to confirm lineage appears correctly - Condensed "On Astro" marketing into a single-line note **testing-dags, tracing-downstream-lineage, troubleshooting-astro-deployments (+5% each)** - Tightened verbose instruction blocks - Removed intro prose that restated the skill title </details> Honest disclosure — I work at @tesslio where we build tooling around skills like these. Not a pitch - just saw room for improvement and wanted to contribute. Want to self-improve your skills? Just point your agent (Claude Code, Codex, etc.) at [this Tessl guide](https://docs.tessl.io/evaluate/optimize-a-skill-using-best-practices) and ask it to optimize your skill. Ping me - [@popey](https://github.com/popey) - if you hit any snags. Thanks in advance 🙏
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

1 participant