.agents/skills/contributor-guide/SKILL.md
How to contribute, build, test
npx skillsauth add ronniegeraghty/hyoka contributor-guideInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Hyoka is an open evaluation tool for Azure SDK code generation. Contributors should follow the setup and workflow documented here to ensure smooth development and testing.
go version)# Clone the repo
git clone https://github.com/ronniegeraghty/hyoka.git
cd hyoka
# Install dependencies (uses go.work for workspace)
go mod download
# Build the CLI
go build ./hyoka/...
# Run the CLI
go run ./hyoka <command>
Create a branch from ronniegeraghty/dev:
git checkout ronniegeraghty/dev
git pull origin ronniegeraghty/dev
git checkout -b ronniegeraghty/issue-{N}-{description}
Configure git identity:
git config user.name "ronniegeraghty"
git config user.email "[email protected]"
hyoka/internal/ or hyoka/cmd/go test -race ./hyoka/...
go fmt ./hyoka/...
go run ./hyoka run --prompt-id key-vault-dp-python-crud \
--config baseline/claude-opus-4.6 --log-level debug
Ensure tests pass with race detector:
go test -race ./hyoka/...
Clean up orphaned sessions from test runs:
go run ./hyoka clean
Include the Co-authored-by trailer in commit message:
git commit -m "Fix eval engine timeout handling
- Add timeout context to generation phase
- Capture timeout error in action timeline
Closes #162
Co-authored-by: Copilot <[email protected]>"
Push to your fork:
gh auth switch --user ronniegeraghty
git push origin ronniegeraghty/issue-{N}-{description}
Create a PR against ronniegeraghty/dev:
gh pr create --base ronniegeraghty/dev \
--title "Fix: eval engine timeout" \
--body "Closes #162"
Link to board: Edit issue on Azure/projects/424 and set Status → "In Progress"
For iterating on changes, run a single prompt × single config (fastest):
go run ./hyoka run --prompt-id key-vault-dp-python-crud \
--config baseline/claude-opus-4.6 --log-level debug
Python prompts finish quickest (5-10 minutes). After each run, clean up:
go run ./hyoka clean
# View the debug log
cat hyoka-debug.log
# Search for role-prefixed output
grep "role=" hyoka-debug.log | head -20
go run ./hyoka serve
# Open http://localhost:8080
hyoka/internal/graders/my_grader.gotype MyGrader struct { /* fields */ }
func (g *MyGrader) Kind() string { return "my" }
func (g *MyGrader) Grade(ctx context.Context, input GraderInput) (GraderResult, error) { /* ... */ }
hyoka/internal/graders/registry.gohyoka/internal/graders/my_grader_test.goconfigs/*.yamlhyoka/cmd/mycommand.govar myCmd = &cobra.Command{
Use: "mycommand",
RunE: func(cmd *cobra.Command, args []string) error { /* ... */ },
}
func init() { rootCmd.AddCommand(myCmd) }
hyoka/cmd/mycommand_test.godocs/architecture.mddocs/cli-reference.mddocs/contributing.mdhyoka/README.mdCloses #{issue} in description(Handled by maintainers)
main branchmain branch (use feature branches)go test -racehyoka clean)development
Identifies Azure SDK packages in generated code and checks whether they are the latest available versions. Use during code review to catch outdated dependencies.
development
Sets up build environments for generated Azure SDK code samples and attempts to compile/build without modifying generated files. Use during review to verify code compiles correctly.
development
# Java SDK Validation Skill You are a **Java Azure SDK validation reviewer** for generated code samples. Your job is to check whether generated Java code follows modern Azure SDK for Java conventions and flag violations of common anti-patterns that LLMs frequently produce. ## Rules 1. **NEVER modify generated code.** You are evaluating, not fixing. 2. Report all findings honestly — pass or fail with specific evidence. 3. Check every rule below. A single violation in a category means that cate
development
Reads generated Azure SDK code files and adds inline review comments without changing any actual code. Use during code review to annotate quality issues, best practices, and suggestions.