Skip to content

feat: improve skill quality scores for MetaClaw#36

Open
rohan-tessl wants to merge 1 commit intoaiming-lab:mainfrom
rohan-tessl:improve/skill-review-optimization
Open

feat: improve skill quality scores for MetaClaw#36
rohan-tessl wants to merge 1 commit intoaiming-lab:mainfrom
rohan-tessl:improve/skill-review-optimization

Conversation

@rohan-tessl
Copy link

Hey πŸ‘‹ @huaxiuyao

I ran your skills through tessl skill review at work and found some targeted improvements.

image

Here's the full before/after:

Skill Before After Change
do-not-retry-without-diagnosis 43% 86% +43%
tool-selection-strategy 51% 85% +34%
source-evaluation 53% 86% +33%
structured-step-by-step-reasoning 51% 84% +33%
async-communication-etiquette 61% 93% +32%
idempotent-script-design 61% 86% +25%
plan-before-multi-step-execution 51% 76% +25%
structured-progress-update 70% 88% +18%
technical-writing-clarity 75% 93% +18%
uncertainty-acknowledgment 70% 88% +18%
graceful-error-recovery 76% 93% +17%
audience-aware-communication 59% 76% +17%
secure-code-review 70% 86% +16%
prioritize-by-impact 70% 81% +11%
structured-research-workflow 75% 86% +11%
structured-logging-and-observability 77% 86% +9%
task-decomposition 76% 84% +8%
auth-and-authorization-patterns 81% 86% +5%
secrets-management 81% 86% +5%

Changes made:

  • Expanded frontmatter descriptions with explicit "Use when..." trigger clauses, natural user-facing keywords, and concrete action lists
  • Removed unknown category frontmatter key causing validation warnings
  • Preserved all domain expertise and specialized terminology
  • Kept body content changes minimal β€” existing skill bodies were already well-structured

Skills left unchanged: verify-before-irreversible-action (already at 93%), sql-best-practices, debug-systematically, context-window-management, agent-task-handoff (already scoring well)

Honest disclosure β€” 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 https://docs.tessl.io/evaluate/optimize-a-skill-using-best-practices and ask it to optimize your skill. Ping me - @rohan-tessl - if you hit any snags.

Thanks in advance πŸ™

Hey πŸ‘‹ @huaxiuyao

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 |
|-------|--------|-------|--------|
| do-not-retry-without-diagnosis | 43% | 86% | +43% |
| avoid-hallucinating-specifics | 48% | 86% | +38% |
| avoid-acting-on-assumptions | 43% | 80% | +37% |
| clarify-ambiguous-requests | 51% | 88% | +37% |
| tool-selection-strategy | 51% | 85% | +34% |
| source-evaluation | 53% | 86% | +33% |
| structured-step-by-step-reasoning | 51% | 84% | +33% |
| test-before-ship | 53% | 86% | +33% |
| async-communication-etiquette | 61% | 93% | +32% |
| data-validation-first | 51% | 77% | +26% |
| idempotent-script-design | 61% | 86% | +25% |
| plan-before-multi-step-execution | 51% | 76% | +25% |
| input-validation-and-sanitization | 70% | 93% | +23% |
| avoid-scope-creep | 59% | 80% | +21% |
| structured-progress-update | 70% | 88% | +18% |
| technical-writing-clarity | 75% | 93% | +18% |
| uncertainty-acknowledgment | 70% | 88% | +18% |
| graceful-error-recovery | 76% | 93% | +17% |
| audience-aware-communication | 59% | 76% | +17% |
| robust-error-handling-in-scripts | 77% | 94% | +17% |
| visualization-selection | 76% | 93% | +17% |
| secure-code-review | 70% | 86% | +16% |
| prioritize-by-impact | 70% | 81% | +11% |
| structured-research-workflow | 75% | 86% | +11% |
| professional-email-drafting | 84% | 93% | +9% |
| structured-logging-and-observability | 77% | 86% | +9% |
| codebase-navigation | 76% | 84% | +8% |
| task-decomposition | 76% | 84% | +8% |
| auth-and-authorization-patterns | 81% | 86% | +5% |
| git-workflow | 81% | 86% | +5% |
| secrets-management | 81% | 86% | +5% |

Changes made:

- Expanded frontmatter descriptions with explicit "Use when..." trigger clauses,
  natural user-facing keywords, and concrete action lists
- Removed unknown `category` frontmatter key causing validation warnings
- Preserved all domain expertise and specialized terminology
- Kept body content changes minimal β€” existing skill bodies were already
  well-structured

Skills left unchanged: verify-before-irreversible-action (already at 93%),
sql-best-practices, debug-systematically, context-window-management,
agent-task-handoff (already scoring well)

Honest disclosure β€” 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 https://docs.tessl.io/evaluate/optimize-a-skill-using-best-practices
and ask it to optimize your skill. Ping me - @rohan-tessl - if you
hit any snags.

Thanks in advance πŸ™
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant