Atlassian Forge · Codegeist 2025
Smooth Operation
An Atlassian Forge app that turns Jira and Confluence chaos into clarity. Real-time heatmaps, analytics, smart search, and vulnerability scanning help teams spot blockers, manage risk, and move faster.
4 features
Heatmap, analytics, smart search, vulnerability scan
Privacy first
Runs entirely inside the Forge sandbox, no external AI APIs
The problem
Standard Jira boards hide context. Teams click through dozens of issues to check deadlines, security status, or dependency chains. The tool should surface risk and context up front, not make people hunt for it.
- Risk: Which issues are near deadline or contain sensitive data?
- Agility: Which items are stalling, and where is the information?
- Backlog: How can the whole backlog be visualized for sprint planning?
What it does
- Work Items Heatmap: Color-coded board based on status and deadline proximity for hassle-free backlog management and sprint planning.
- Advanced Analytics: Activity score, last activity, and sub-task progress for standups, plus due-date actions when dates are missing or overdue.
- Embedded Smart Search: Context-aware search across Jira (JQL) and Confluence (CQL) so people can find related tasks and docs without pinging teammates.
- Vulnerability Scanning: Checks issue content for secrets like API keys and passwords, with one-click Fix (hash) or Remove actions.
How we built it
- Built entirely on Atlassian Forge with React and the native UI Kit for speed, security, and deep product integration.
- Privacy first: Smart search optimization and vulnerability scanning run inside the Forge sandbox with no external AI or API services.
- Heatmap: Dynamic rendering and custom color logic (red through blue) from live Jira status and due dates, with pagination for large backlogs.
- Smart Search: Uses issue context (project key, summary keywords) to enhance queries before hitting JQL and CQL APIs.
- Analytics: Custom metrics drive UI, including conditional due-date buttons and sub-task progress bars.
Challenges
- Scale: Large-backlog heatmap rendering required pagination and chunked loads (6 items at a time).
- Sandboxed scanning: Regex-based secret detection had to stay fast inside Forge resource limits.
- Unified search: Aggregating Jira and Confluence results into one readable pane needed careful data shaping.
Outcome
- Delivered a private "smart" experience without external AI services, suited to enterprise security needs.
- Shipped a heatmap that makes backlog and deadline risk readable at a glance.
- Made security actionable with one-click Fix and Remove for sensitive content.
- Submitted to Codegeist 2025: Atlassian Williams Racing Edition.
Tech
Atlassian Forge
React
Node.js
JavaScript
JQL
Bitbucket
YAML
Full write-up on Devpost
Inspiration, build notes, challenges, and next steps for Smooth Operation.