Tessl Skill Review & Optimize Action

A GitHub Action that reviews SKILL.md files changed in a pull request and optionally optimizes them with AI-powered suggestions.

Usage

Examples below pin this action to a specific commit SHA so your workflow stays reproducible and does not pick up unexpected changes from main. Replace the SHA when you intentionally upgrade; you can also use a release tag if one is published.

Review only (no authentication required)

name: Skill Review
on:
  pull_request:
    paths: ['**/SKILL.md']

jobs:
  review:
    runs-on: ubuntu-latest
    permissions:
      pull-requests: write
      contents: read
    steps:
      - uses: actions/checkout@v4
      - uses: tesslio/skill-review-and-optimize@bff9490027d60847df6494fdac7dccfb3ad82948

Any PR that modifies a SKILL.md file gets an automated review comment with scores and feedback.

Review + Optimize (requires Tessl API token)

- uses: tesslio/skill-review-and-optimize@bff9490027d60847df6494fdac7dccfb3ad82948
  with:
    optimize: true
    tessl-token: ${{ secrets.TESSL_API_TOKEN }}

When optimize is enabled, the action reviews each skill, then runs AI-powered optimization and posts the suggested improved SKILL.md content directly in the PR comment. Users can then comment /apply-optimize to commit the optimized content to the PR branch.

Apply optimized content (via /apply-optimize comment)

Add a second workflow to let PR authors apply suggested optimizations with a single comment:

name: Apply Skill Optimization
on:
  issue_comment:
    types: [created]

jobs:
  apply:
    if: >
      github.event.issue.pull_request &&
      contains(github.event.comment.body, '/apply-optimize')
    runs-on: ubuntu-latest
    permissions:
      pull-requests: write
      contents: write
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0
      - uses: tesslio/skill-review-and-optimize@bff9490027d60847df6494fdac7dccfb3ad82948
        with:
          mode: apply

When a user comments /apply-optimize on a PR that has a Tessl review comment with optimization suggestions, this workflow extracts the optimized content and commits it directly to the PR branch.

Getting a TESSL_API_TOKEN

1. Sign up or log in at tessl.io
2. Get your API token at tessl.io/account/api-keys
3. Add it as a repository secret named TESSL_API_TOKEN

No CLI install or workspace setup required.

Inputs

Input	Description	Default
mode	Action mode: review or apply	review
path	Root path to search for SKILL.md files	.
comment	Whether to post results as a PR comment	true
fail-threshold	Minimum score (0-100) to pass. Set to 0 to never fail.	0
optimize	Run skill optimization after review (requires tessl-token)	false
optimize-iterations	Max optimization iterations (1-10)	3
tessl-token	Tessl API token for optimize mode	(optional)

Setting a quality gate

- uses: tesslio/skill-review-and-optimize@bff9490027d60847df6494fdac7dccfb3ad82948
  with:
    fail-threshold: 70

PRs with any skill scoring below 70% will fail the check.

How it works

1. Detects which SKILL.md files were changed in the PR
2. Installs the Tessl CLI
3. Runs tessl skill review on each changed skill
4. If optimize: true and tessl-token is provided, runs optimization and captures suggested improvements
5. Posts (or updates) a review comment on the PR with scores, feedback, and optimization suggestions
6. Optionally fails the check if any score is below the threshold
7. If a user comments /apply-optimize, the apply workflow extracts optimized content from the review comment and commits it to the PR branch

When optimize is enabled but no token is provided, the action runs review-only and includes a prompt in the comment to set up optimization.

Comment behavior

The action posts a single comment per PR. On subsequent pushes, it updates the existing comment rather than creating a new one. Optimized skills show before/after score badges and the suggested content in a collapsible section.

Local development

bun install
bun run lint
bun test

License

MIT