JCMA Analyzer
Upload a JCMA export, analyze it entirely in your browser, and generate a report.
Waiting for a JCMA ZIP file.
AI Provider
Keys are saved in your browser only — never sent anywhere except directly to the AI provider via this app's server function. Cloudflare environment secrets take precedence if set.
AI readiness: checking...
Knowledge Frameworks
Framework context is injected into every AI analysis.
Praecipio Intelligence Framework
Default · Always active
Framework Log
Updated after each AI analysis run.
No analysis run yet.
Help
Frameworks are GitHub repos containing Praecipio methodology files. The default Praecipio Intelligence Framework is always injected. Additional frameworks require a GitHub PAT with repo:read access.
Framework log shows which files were fetched per analysis, their size, approximate token count, and whether they were served from the 1-hour cache.
Token limits — Anthropic Claude Sonnet 4.6: 1M input / 64k output. Gemini 2.5 Flash: 1M input / 64k output. Current max output is set to 8,192 tokens.
Providers — Set ANTHROPIC_API_KEY for Claude or GEMINI_API_KEY for Gemini in your Cloudflare Pages environment secrets. Both can be set; Auto mode picks the first available.
Output
Migration Assessment
Core metrics
Derived from the JCMA export archive.
Top projects by issue count
Uses mapping file if provided.
| Project | Issue count |
|---|
Detected file inventory
Required files should be present for best accuracy.