AI-first linting
Column-aware COBOL linting, dependency tracing, and modernization tips surface in seconds instead of long review cycles.
The AI Platform for
$npm install -g zcrafter
$export AI_API_KEY="skh-..."
$zcrafter
Install Zcrafter, set your API key, and start the CLI.
Used by engineers at
Why teams choose Zcrafter
Everything you need to keep mainframe delivery moving
Zcrafter combines AI-assisted linting, JCL automation, and embedded knowledge so COBOL teams can ship confidently without leaving the terminal.
Watch mode automation
Background watchers keep datasets clean by proposing fixes, documenting rationale, and waiting for approvals before write back.
JCL you can trust
Submit, monitor, and analyze jobs with contextual abend explanations so operators resolve incidents faster.
Dataset intelligence
Catalogs members, flags duplicates, and keeps hygiene metrics so large COBOL libraries stay maintainable.
Smart connections
Automatic port discovery, protocol negotiation, and intelligent error handling keep z/OSMF sessions stable.
Bring your own keys
Use OpenRouter or vendor-specific API keys. Billing and quotas stay under your control.
Privacy First
Built to keep your code on your machine
No code storage
Commands execute locally; we never copy or cache your datasets, source, or job output.
Explicit consent
Every external request requires a signed token and user approval so nothing leaves your terminal unexpectedly.
Auditable binaries
Signed releases with published checksums let teams verify integrity before rollout.
Operations Benchmark
Measured on core mainframe tasks
We compared Zcrafter-assisted workflows with traditional TSO and Zowe routines across linting, dataset hygiene, and JCL troubleshooting.
| Function | Zowe / TSO | Zcrafter |
|---|---|---|
| COBOL linting | Manual syntax checks, copybook lookups take ~12 min per module | Automated lint run averages ~4 min with inline suggestions |
| Dataset hygiene | Custom REXX scripts and manual reviews | Prebuilt hygiene pack flags duplicates and dead paragraphs |
| JCL troubleshooting | Manual JES spool inspection with cross references | Automated abend summaries with remediation pointers |
| Documentation lookup | PDF search plus TSO help panels | Contextual references surfaced alongside CLI guidance |
| Team onboarding | Manual write-ups and shadow sessions | Interactive command tips and checklists embedded in CLI |
| Security posture | Local scripts with inconsistent review | Signed CLI, optional approval gates, zero code retention |
FAQ
Everything you need to know before you get started
About Us
Empowering Mainframe Developers
Zcrafter is an AI-powered toolkit designed to streamline mainframe development workflows. We combine artificial intelligence with mainframe expertise to help developers work more efficiently with COBOL, JCL, and legacy systems. Our mission is to bridge the gap between modern development practices and traditional mainframe environments, making it easier for teams to maintain and modernize critical enterprise systems.
Keep exploring Zcrafter
Dive into research, review our privacy posture, or follow the quickstart docs to see Zcrafter CLI in action.