Skip to content

feat: improve vale-linter skill score from 82% to 97%#10454

Open
yogesh-tessl wants to merge 1 commit into
circleci:mainfrom
yogesh-tessl:improve/skill-review-optimization
Open

feat: improve vale-linter skill score from 82% to 97%#10454
yogesh-tessl wants to merge 1 commit into
circleci:mainfrom
yogesh-tessl:improve/skill-review-optimization

Conversation

@yogesh-tessl

Copy link
Copy Markdown

Hey @rosieyohannan 👋

the way the docs are structured with Antora and component-based architecture makes the content really navigable. Having the contributing guide and community links front and center shows you want people involved, not just reading.

ran your skills through tessl skill review at work and found some targeted improvements for your vale-linter skill. Here's the before/after:

Skill Before After Change
vale-linter 82% 97% +15%
Changes made
  • Extracted templates - moved commit message, PR body, and summary report templates into a new TEMPLATES.md bundle file for progressive disclosure
  • Removed "Common Vale Errors" section - Claude already knows how to fix passive voice, Oxford commas, hedging, contractions, and gender-neutral language without being told
  • Removed redundant "When to Use" section - the frontmatter description already provides comprehensive trigger terms
  • Condensed severity explanation - replaced a 6-line breakdown of Vale severity levels with a single actionable line
  • Trimmed "Best Practices" and "Notes" sections - these repeated guidance already present in the workflow steps
  • Preserved all CircleCI domain knowledge - .vale.ini, styles/circleci-docs/, AGENTS.md references, AsciiDoc formatting rules, and branch naming conventions all retained
  • Ensured description uses quoted string format in frontmatter

also stress-tested your vale-linter skill against a few real-world task evals and it held up really well on batch processing AsciiDoc files with mixed error severities across nested directory structures. Kudos for that.

quick honest disclosure. I work at https://github.com/tesslio where we build tooling around skills like these. Not a pitch, just saw room for improvement and wanted to contribute.

if you want to self-improve your skills, or define your own scenarios to pressure test, just ask your agent (Claude Code, Codex, etc.) to evaluate and optimize your skill with Tessl. Ping me @yogesh-tessl, if you hit any snags.

Hey @rosieyohannan 👋

I ran your skills through `tessl skill review` at work and found some targeted improvements for your `vale-linter` skill. Here's the full before/after:

| Skill | Before | After | Change |
|-------|--------|-------|--------|
| vale-linter | 82% | 97% | +15% |
| content-review | 67% | 67% | — |
| monthly-docs-report | 86% | 86% | — |

# Description

Optimized the `vale-linter` skill to be leaner and more token-efficient while preserving all domain-specific behavior. The skill went from 245 lines to 66 lines — same capabilities, significantly less noise for the agent at runtime.

# Reasons

The `vale-linter` skill scored 82% on review, with the content dimension at 55%. The main issues: inline templates bloating the file, a "Common Vale Errors" section explaining things Claude already knows (passive voice, Oxford commas, contractions), and a "When to Use" section duplicating the frontmatter triggers.

<details>
<summary>Changes made</summary>

- **Extracted templates** — moved commit message, PR body, and summary report templates into a new `TEMPLATES.md` bundle file for progressive disclosure
- **Removed "Common Vale Errors" section** — Claude already knows how to fix passive voice, Oxford commas, hedging, contractions, and gender-neutral language without being told
- **Removed redundant "When to Use" section** — the frontmatter description already provides comprehensive trigger terms
- **Condensed severity explanation** — replaced a 6-line breakdown of Vale severity levels with a single actionable line
- **Trimmed "Best Practices" and "Notes" sections** — these repeated guidance already present in the workflow steps
- **Preserved all CircleCI domain knowledge** — `.vale.ini`, `styles/circleci-docs/`, `AGENTS.md` references, AsciiDoc formatting rules, and branch naming conventions all retained
- **Ensured description uses quoted string format** in frontmatter

</details>

I also stress-tested your `vale-linter` skill against a few real-world task evals and it held up really well on batch processing AsciiDoc files with mixed error severities across nested directory structures. Kudos for that.

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 — [@yogesh-tessl](https://github.com/yogesh-tessl) — if you hit any snags.

Thanks in advance 🙏

**Preview your changes:**
- [ ] View the Vale linter results, select the `ci/circleci: lint` job at the bottom of your PR. You will be redirected to the `vale/lint` job output in CircleCI.
- [ ] Preview your changes, select the `ci/circleci: build` job at the bottom of your PR and you will be redirected to CircleCI. Select the Artifacts tab and select `index.html` to open a preview version of the docs site built for your latest commit.
@yogesh-tessl yogesh-tessl requested review from a team as code owners June 5, 2026 10:04
@yogesh-tessl

Copy link
Copy Markdown
Author

hey @rosieyohannan, just checking in on this one, happy to make any changes if something needs tweaking!

Take your time, totally understand if reviews follow a cadence.

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.

2 participants