- Home
- Customer Journey Mapping Software
- CX Impact
CX Impact
What it is
CX Impact is a portable skill for AI agents that turns supplied documents, process descriptions, and numbers into visual maps. It can create customer journey maps, service blueprints, process maps, AS IS / TO BE comparisons, and sales funnels. Its intended users include product managers, analysts, designers, and service teams. Results include interactive HTML, editable JSON, and a source snapshot; the viewer exports SVG, while funnels can also export CSV. Generated HTML opens without a server or CX Impact account. Installation is documented for Codex, Claude Code, OpenCode, and DeepSeek Harness. Map creation happens in the user's agent environment with the selected model provider, and the generated HTML itself makes no network requests. Requirements, observations, hypotheses, and open questions are kept distinct, with users advised to review sources and unknowns. Quantitative funnels require user-provided data and metric definitions; missing values remain unknown. The viewer has no visual editor, so revisions use the prior JSON and new information provided to the agent. The MIT-licensed project code is free; AI agent and model usage depend on the user's plan.
For and against
For
- Creates five types of maps and comparisons
- Exports interactive HTML, editable JSON, and a source snapshot
- Generated HTML works without a server or account
- Map creation stays in the user's agent environment
- MIT-licensed project code is free
Against
- Quantitative funnels require supplied data and metric definitions
- Viewer has no visual editor
- Renderer needs Node.js 18+
- AI agent and model usage depend on the user's plan
The verdict
CX Impact turns supplied information into several map types and provides multiple output formats. Users should expect to revise maps through the agent rather than a visual editor, and quantitative funnels depend on supplied metrics.
Who it suits
CX Impact suits product managers, analysts, designers, and service teams who want an AI agent to turn supplied materials into visual maps. It may also suit users who want editable JSON or shareable HTML output.
Plans and prices
Open-source project
MIT-licensed code · AI-agent/model usage depends on chosen provider plan · consulting separate
Read 15 Oct 2026 · maxshelukhanov.ru
CX Impact
MIT-licensed project code · AI agent and model usage depend on your plan
Read 4 Oct 2026 · maxshelukhanov.ru
On the record
- Agent integrations
- Installation is documented for Codex, Claude Code, OpenCode and DeepSeek Harness.maxshelukhanov.ru · read 15 Oct 2026
- API and SDKs
- The published integration mechanism is an agent skill with JSON inputs and outputs; no separate API or SDK is documented on the product guide or README.github.com · read 15 Oct 2026
- AS IS / TO BE
- The tool compares current and target scenarios and keeps correspondences and sources visible.maxshelukhanov.ru · read 15 Oct 2026
- Audience
- The maker identifies product managers, analysts, designers, and service teams as its intended users.maxshelukhanov.ru · read 4 Oct 2026
- Browser verification
- Chrome was checked at 1366×900 and 1920×1080; other browser engines and screen readers are not claimed as tested.github.com · read 15 Oct 2026
- Codex invocation
- Codex users invoke the skill with `$cx-impact`.github.com · read 15 Oct 2026
- Customer journey maps
- The customer-journey view describes user goals, interactions and barriers identified in supplied materials.maxshelukhanov.ru · read 15 Oct 2026
- Data processing location
- Map creation runs in the user's agent environment through the selected model provider.maxshelukhanov.ru · read 15 Oct 2026
- Documentation depth
- The repository links installation, troubleshooting, format, verification, evidence, contribution and changelog documentation.github.com · read 15 Oct 2026
- Editing
- The viewer has no visual editor; users can revise results by providing the previous JSON and new information to the agent.maxshelukhanov.ru · read 4 Oct 2026
- Editor limitation
- The viewer is not a visual editor and is not a complete BPMN tool.github.com · read 15 Oct 2026
- Evidence handling
- Requirements, observations, hypotheses, proposals and unknowns remain separate in generated maps.github.com · read 15 Oct 2026
- Evidence limitation
- Automated checks cannot independently verify person identity or cohort maturity, and human source review remains necessary.github.com · read 15 Oct 2026
- Experimental status
- The README describes CX Impact as an experimental public product intended to structure discussion rather than prove authored claims.github.com · read 15 Oct 2026
- Funnel limits
- A quantitative funnel requires user-supplied data and metric definitions; missing values remain unknown.maxshelukhanov.ru · read 4 Oct 2026
- Funnel size limit
- Funnels support up to 32 transition nodes and 48 edges, so larger scenarios may need multiple maps.github.com · read 15 Oct 2026
- Installation command
- The documented installer command is `npx skills add shelasmax/cx-impact --skill cx-impact`.github.com · read 15 Oct 2026
- Installer requirement
- The verified Skills CLI installer requires Node.js 22.20 or newer, npm and internet access.maxshelukhanov.ru · read 15 Oct 2026
