ronniegeraghty/report-generation
.agents/skills/report-generation/SKILL.md
JSON/HTML/Markdown reports, template system
$ install --global
skillsauthnpx skillsauth add ronniegeraghty/hyoka report-generationInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 16, 2026, 5:47 PM6.1s1 file scanned
SKILL.md
- name:
- report-generation
- description:
- JSON/HTML/Markdown reports, template system
- domain:
- architecture
- confidence:
- high
- source:
- hyoka/internal/report/*.go
Context
Hyoka generates multi-format evaluation reports (JSON, HTML, Markdown) from a unified internal representation. Reports capture the entire evaluation timeline: generation, build, grading, and review phases.
Report Formats
JSON Report
The canonical format — contains all raw data and is used to generate other formats:
{
"metadata": {
"run_id": "eval-2024-04-06-123456",
"timestamp": "2024-04-06T12:34:56Z",
"prompt": {
"id": "identity-dp-python-default-credential",
"service": "identity",
"language": "python"
},
"config": {
"name": "baseline/claude-opus-4.6"
}
},
"generation": {
"status": "success",
"duration_ms": 45000,
"code": "# Generated Python code...",
"action_timeline": {
"events": [...],
"summary": {...}
}
},
"build": {
"status": "success",
"duration_ms": 5000,
"stdout": "Compilation successful",
"stderr": ""
},
"graders": [
{
"kind": "behavior",
"name": "tool_compliance",
"pass": true,
"score": 1.0,
"message": "All required tools used"
}
],
"review": {
"status": "success",
"reviewers": [
{
"model": "claude-opus-4.6",
"score": 0.9,
"findings": "Code is mostly correct..."
}
],
"consensus_score": 0.85
}
}
HTML Report
Browser-viewable report with interactive elements:
<!DOCTYPE html>
<html>
<head>
<title>Evaluation Report: identity-dp-python</title>
<style>...</style>
</head>
<body>
<div class="report-container">
<h1>Evaluation Report</h1>
<section class="generation">
<h2>Generation Phase</h2>
<code>{{ .Code }}</code>
<div class="timeline">...</div>
</section>
<section class="graders">
<h2>Grading Results</h2>
...
</section>
</div>
<script>
// Interactive features (expand/collapse, filtering)
</script>
</body>
</html>
Markdown Report
Human-readable text format for documentation and sharing:
# Evaluation Report: identity-dp-python
## Generation
- Status: Success
- Duration: 45s
## Code
\`\`\`python
# Generated code...
\`\`\`
## Graders
| Grader | Status | Score | Message |
|--------|--------|-------|---------|
| behavior | Pass | 1.0 | All required tools used |
| lint | Pass | 0.95 | 1 warning |
## Review Panel
- Claude Opus 4.6: 0.9
- GPT-5.4: 0.85
- Consensus: 0.875
Report Generation Pipeline
- Evaluation Completion: Engine returns
EvaluationResultwith all phase outputs - JSON Marshaling: Convert result to JSON bytes
- File Write: Save JSON to
reports/{run_id}/report.json - Template Rendering: Apply HTML/Markdown templates to JSON
- Asset Linking: Copy static assets (CSS, JS) to report directory
Template System
Reports use Go's text/template package:
// report.go
type ReportTemplate struct {
HTML *template.Template
Markdown *template.Template
}
func (rt *ReportTemplate) RenderHTML(data EvaluationResult) (string, error) {
var buf bytes.Buffer
if err := rt.HTML.Execute(&buf, data); err != nil {
return "", err
}
return buf.String(), nil
}
Template Variables
Templates can access all fields from EvaluationResult:
{{- .Metadata.RunID }}
{{- .Generation.Status }}
{{- range .Graders }}
{{- .Kind }}: {{- .Score }}
{{- end }}
Report Storage
Reports are written to:
reports/
{run_id}/
report.json # Canonical data
report.html # Browser view
report.md # Text format
assets/
style.css
script.js
Multi-Report Aggregation
For batch runs (multiple prompts × configs):
{
"batch_metadata": {
"run_date": "2024-04-06",
"prompts": 5,
"configs": 2,
"total_evals": 10
},
"summary": {
"avg_score": 0.82,
"pass_count": 8,
"fail_count": 2
},
"reports": [
{ "prompt_id": "...", "config": "...", "result": {...} }
]
}
Re-rendering from JSON
The rerender command regenerates HTML/Markdown from saved JSON:
go run ./hyoka rerender --report-id eval-2024-04-06-123456 --format html
This allows updating templates without re-running evaluations.
Code Locations
- Report structure:
hyoka/internal/report/report.go - JSON marshaling:
hyoka/internal/report/json.go - HTML/Markdown rendering:
hyoka/internal/report/render.go - Templates:
hyoka/templates/
Anti-Patterns
- Hardcoding paths in templates (use config for asset paths)
- Not including action timeline in JSON (required for reproducibility)
- Rendering HTML/Markdown before generation completes (wait for all phases)
- Losing build/grader details in summary format (preserve full data in JSON)
- Assuming report directory exists (create it before writing)
Related Skills
ronniegeraghty/sdk-version-check
development
VerifiedTrustedCommunity
Identifies Azure SDK packages in generated code and checks whether they are the latest available versions. Use during code review to catch outdated dependencies.
1SKILL.mdUpdated Apr 16, 2026
ronniegeraghty/reviewer-build
development
VerifiedTrustedCommunity
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.
1SKILL.mdUpdated Apr 16, 2026
ronniegeraghty/skills/reviewer/java-sdk-validation
development
VerifiedTrustedCommunity
# 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
1SKILL.mdUpdated Apr 16, 2026
ronniegeraghty/code-review-comments
development
VerifiedTrustedCommunity
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.
1SKILL.mdUpdated Apr 16, 2026