What's ready. And what isn't.
Every JuhJuh feature carries a maturity level. 53 stable, 31 preview, 2 experimental. Published, not implied.
Stable
53
In production, exercised daily, on by default. Depend on it.
Preview
31
Built and usable, but off by default or not yet broadly exercised. Turn it on deliberately.
Experimental
2
Early. Expect it to change shape. Do not build a process on it yet.
These levels are our assessment, not test-derived. A feature is “stable” because it runs in production and is exercised daily — not because a named test suite proves it. We would rather tell you that than imply evidence we do not have.
A feature moves up on evidence and down the moment it stops earning its level. If something here is wrong, tell us and we will change it.
AI agents at work
- Intelligent Ticket Analysis JuhJuh classifies complexity, identifies affected files, and generates structured PRDs before a single line of code is written. Teams stop debating scope and start building. Stable
- AI Triage and Priority Routing Every incoming ticket is classified by type, priority, and best-fit team. Tickets flow to the right people without manual sorting. Stable
- Prompt Engineering and Management Generate sequenced execution prompts from analyzed tickets. Template library with include directives, placeholder substitution, and version control. Stable
- Agent Runs You Can Watch Agents work on their own computer with sandboxed configuration and health monitoring, and you watch the output as it happens. Every run tracks tokens, cost, and duration. Stable
- Autonomous AI Agents Conversational AI agents with autonomous reasoning, 200+ tool integrations across 23 categories, 11 specialist agents, trigger rules, skill tracking, and anti-hallucination safeguards. Stable
- Parallel Prompt Execution Dependency-based parallel execution runs prompts concurrently when no unmet dependencies exist. Runner reservation manages parallel slots automatically. Stable
- Idea Capture and AI Expansion A lightweight idea inbox lets anyone capture a thought in seconds. AI transforms a one-liner into a structured specification with requirements and edge cases. Preview
- AI Ticket Rewriting Refactor poorly written tickets into clear, actionable specifications with one click. Improves first-pass success rates and reduces back-and-forth. Preview
Project Management
- Multi-Project Dashboard Project gallery with live stats, resource connections, and classification badges. Color-coded cards with custom icons across all your codebases. Stable
- Kanban Board with Real-Time Updates 15 ticket statuses across a live-updating Kanban board. Instant updates, type filtering, quality indicators, and keyboard shortcuts. Stable
- Ticket Hierarchy and Epic Management Epics contain stories, stories contain subtasks, and progress rolls up automatically. Auto-Play executes child tickets sequentially with shared context. Stable
- Visual Workflow Graph Interactive D3.js workflow graph with drag-and-drop, zoom, pan, minimap, search, and fullscreen. Visualize ticket dependencies and execution flow at a glance. Preview
- Automatic Ticket Decomposition Oversized tickets split automatically when prompts exceed size limits. Recursive epic handling ensures grandparent chains resolve cleanly. Preview
- 9 Project Templates IoT Platform, Microservices, SaaS Application, Marketing Campaign, and more. Pre-configured classification, settings, and resource types. Stable
- Roadmap and Timeline View Epics and stories on an interactive timeline with target dates and progress tracking. Generate shareable public links for stakeholders. Preview
- Global and Project-Scoped Search Full-text search spans tickets, ideas, knowledge entries, and resources. Scoped to a single project or across the entire platform. Stable
Code across repositories
- Managed Repository Workspace Clone and sync several repositories in one workspace. Stable
- Service Dependency Mapping Visualize how microservices connect with 7 dependency types. Preview
- Branch Isolation Every execution runs in an isolated branch. Main branch is never at risk. Stable
Integrations
- Issue Tracker Sync Two-way sync with Jira (full), Linear, and GitHub Issues. Stable
- Git Platform Support Native PR creation and pipeline monitoring for Bitbucket, GitHub, and GitLab. Stable
- Bidirectional Slack Integration Message sync, ticket creation via emoji reactions, link unfurl with Block Kit cards, thread mirroring, intent-classified message routing, and per-channel AI guidance. Stable
- Notification Channels Slack blocks, Discord embeds, Teams Adaptive Cards, and HTML email. Stable
- Infrastructure Connections SSH status checks, Kubernetes deployments, and custom webhooks. Stable
- Error Log Monitoring Detect production errors, create tickets, and generate fixes automatically. Preview
Safety and Compliance
- Five-Layer Safety Model Prompt review, branch isolation, diff review, PR approval, and pipeline gate. Stable
- IP allowlisting for self-hosted deployments Limit access to the IP ranges you choose, with proxy-aware trust rules. Stable
- Rate Limiting Sliding-window rate limits on sensitive endpoints. Stable
- Full Audit Trail Who, what, when, and where for every action. Stable
- Environment Variable Sanitization AI agents receive only approved environment variables. Stable
- Role-Based Access Control Owner, Admin, Member, and Viewer across projects and organizations. Stable
Real-Time Collaboration
- Live Updates Instant updates across dashboards, tickets, projects, and executions. Stable
- Live Execution Streaming Watch AI write code line by line in real time. Stable
- Offline-Durable Messaging Messages persist locally before they send and flush automatically on reconnect. After a drop the client gap-fills by cursor, so nothing is missed and no refresh is needed. Stable
- Project-Scoped Broadcasting Events reach only the people working on that project. Stable
Knowledge Management
- Hybrid Vector Search 8 scoring signals combine full-text ranking, semantic similarity, and graph-based relevance. Results ranked by meaning, not just keywords. Stable
- Knowledge Graph Cross-references between knowledge entries with graph-based ranking. Near-duplicate detection and automated fact extraction surface connections across your codebase. Preview
- 14 Knowledge Categories Architecture, domain, patterns, API endpoints, model details, services, business rules, workflows, integrations, configuration, pages, and more. Full-text search across all categories. Stable
- Automatic Knowledge Extraction Automated code analysis and fact extraction sync documentation into categorized entries. CLI integration for local development. Preview
- Agent Instructions Hierarchy Organization, project, team, and user levels with cascading inheritance. Stable
- Lessons From Every Run Capture learnings from completed tickets into future runs. Feedback loop tracks usage across executions for continuous improvement. Preview
From request to finished work
- Tickets That Get Done A ticket can end in a report, a page update, a customer reply or a merged pull request, all in one workspace. Stable
- Diff Review and Approval Review every AI-generated change. Approve, reject, or iterate. Stable
- Ticket Reset and Rollback Reset execution, git state, and analysis with one action. Stable
Automation and Scheduling
- Auto-Play Sequential Execution Execute all child tickets under an Epic with shared sibling context. Stable
- Automation Rules Engine Auto-queue, auto-analyze, auto-execute, and auto-PR creation per project. Preview
- Backlog Health and Auto-Grooming Stale ticket detection, priority decay, and AI grooming suggestions. Preview
- Scheduled Task Management Scheduled automation for recurring sync, health checks, and maintenance. Stable
- Automatic Ticket Decomposition Oversized tickets auto-split when prompts exceed context limits. Recursive epic handling resolves grandparent chains. PRD clears from parent after split. Preview
Analytics and Insights
- Permission-Safe Lookup Columns Resolve values across permitted datasets in one batched page pass. Viewers see only source rows and fields they can access. Preview
- Watcher-Triggered Charms Start an eligible charm when new view rows match a saved condition. Every run records its initiating watcher. Preview
- Add-to-Charm Row Actions Send an eligible view row to a permitted charm with execution-time authorization and the row context preserved. Preview
- Token Usage and Cost Tracking Input tokens, output tokens, and USD cost per ticket, per project, per team. Stable
- Velocity and Cycle Time Metrics Throughput charts, bottleneck identification, and trend tracking. Preview
- Execution Success Rates Prompt-level success/failure tracking for continuous improvement. Stable
Customization
- Project Branding Custom colors, Material Symbols icons, and tag sets per project. Stable
- Brand Voice AI 7-step brand voice wizard defines your writing style across 6 core traits and 15 tone contexts. AI enforces your voice in all 75 content-producing prompt templates. Preview
- Custom Ticket Types Define your own types beyond Bug, Feature, Task, Epic, and Story. Stable
- Prompt Template Library Versioned templates with placeholder substitution and include directives. Stable
- Markdown Editor with Live Preview Toolbar, keyboard shortcuts, and server-rendered preview across 18+ fields. Stable
- Dark Mode First-class dark mode across every screen, not an afterthought. Stable
Enterprise Features
- Multi-Tenant Organizations Independent projects, members, and billing per organization. Stable
- Subscription Tiers Standard, Pro, Premium, and Enterprise plans with configurable limits. Stable
- Invitation and Onboarding Token-based invitations with role pre-assignment and expiration. Stable
- Team-Based Skill Routing 14 team categories with auto-routing based on ticket classification. Preview
- Custom Organization Subdomains Each organization gets its own subdomain with real-time updates, cross-subdomain cookie handling, reserved slug protection, and single-org auto-redirect. Stable
- Editor Plugins JetBrains and VS Code plugins with file type detection, JSON schema validation, custom icons, syntax highlighting, and language configuration. Experimental
Execution Sandboxes
- Per-Ticket Brew Lifecycle Sandboxes auto-create on demand when work starts and auto-destroy when the ticket goes idle. No manual provisioning, no leftover state between runs. Stable
- Live Preview URLs Every ticket gets its own secure preview link with certificates issued automatically. Reviewers see the change in production-like conditions before merge. Stable
- Per-Project Commands Dedicated terminal sessions per project with persistent activity, pinned commands, and replayable history. Frequent workflows rerun across sandboxes in one click. Stable
- Distributed Execution Parallel work runs across isolated branches with safe commit ordering. Multiple tickets advance at the same time without stepping on each other's changes. Stable
- Code and PR Review Surface Code review and pull request threads attach to every ticket. Inline comments, diff annotations, and reviewer status stay tied to the work. Stable
- Sandbox Isolation Network, filesystem, and tool restrictions enforced per brew. Each ticket runs in a contained perimeter with no path back to your main branch or shared state. Stable
Observability
- Uptime Monitoring Multi-region checks for HTTP, TCP, and ping endpoints with public status pages and subscriber alerts. Notified the moment a component degrades. Preview
- Error Tracking Grouped errors with full stack traces, breadcrumbs, and an OTLP-compatible ingest. Spot the change that introduced the regression and link it back to the ticket. Preview
- Infrastructure Metrics Full-stack snapshots for containers, databases, and queries. CPU, memory, slow queries, and connection saturation across every environment. Preview
- Application Metrics Custom metrics with alert rules and templates for the patterns teams ship most. Fire on thresholds, anomalies, or absence of signal. Preview
- Distributed Tracing Span-level traces with automatic service-dependency mapping. Follow a single request across services and pinpoint the slow hop. Preview
- Real User Monitoring Session replay, web vitals, and frontend errors from real browsers. See exactly what users experienced when something broke. Experimental
- Logs Container logs and gateway request logging, searchable across the stack. Correlate logs with traces, metrics, and incidents from a single timeline. Preview
Incident Management
- Incident Lifecycle Timeline entries capture every action, status change, and decision with full context preserved. Replay the incident exactly as it happened. Preview
- On-Call Schedules Rotations, overrides, and escalation policies with timezone-aware handoffs. The right person gets paged, every time, without manual coordination. Preview
- Alert-Driven Incidents Auto-create incidents from alerts using templates that pre-populate severity, affected components, and runbook links. Responders start with context, not a blank page. Preview
- Post-Mortems Structured templates with action-item tracking, owners, and due dates. Lessons turn into tickets that actually get scheduled and shipped. Preview
- AI-Powered Analysis Root-cause hints and anomaly correlation across metrics, logs, and recent deploys. Accelerates triage when every minute counts. Preview
- Public Status Pages Component-level state with subscriber updates and incident history. Communicate honestly without standing up a separate status tool. Preview
Point it at one ticket.
Not a migration. One ticket, one branch, one diff you review before it merges.
