Inspiration
Modern engineering organizations manage thousands of distributed assets across Jira tickets, Confluence technical specifications, code repositories, and runtime infrastructure. When incident escalations occur, teams frequently suffer from severe context fragmentation: the architecture documented in Confluence deviates from the actual Jira backlog and live deployment topology. SentinelForge was conceived to bridge cross-product organizational intelligence seamlessly into an automated sentinel powered by Atlassian Forge and the Model Context Protocol (MCP).
What it does
SentinelForge acts as an intelligent sentinel and Rovo agent within Atlassian Cloud.
- Cross-Product Context Graph: Leverages Atlassian Teamwork Graph to continuously map relationships between Jira issues, Confluence documentation, and system services.
- Drift & Bottleneck Detection: Automatically analyzes dependencies across sprints and alerts teams inside Jira issue panels when architectural or compliance drift is detected.
- Model Context Protocol (MCP) Integration: Connects Atlassian Rovo agents with standard MCP tooling, enabling natural-language investigation of live services and secure automated remediations.
How we built it
- Atlassian Forge: Developed using Forge Functions and declarative modules (jira:issuePanel, rovo:agent).
- Teamwork Graph: Graph traversals to uncover hidden bottlenecks and link distributed project deliverables.
- Node.js Runtime & Forge API: Secure invocation within the Atlassian cloud boundary adhering to strict enterprise scopes.
Challenges we ran into
Designing a performant query model for cross-product graph traversal while honoring strict least-privilege security boundaries was challenging. We resolved this by relying on Forge-native serverless functions and declarative manifest scopes.
Accomplishments that were proud of
- Seamless embedding directly inside the Jira issue view without third-party hosting dependencies.
- Zero-trust architecture: all compute and state remain strictly inside the Atlassian Forge secure sandbox.
- Interoperability with Model Context Protocol standards.
What we learned
Building with Atlassian Forge drastically reduces operational overhead by abstracting auth, storage, and deployment infrastructure. The new Rovo agent capabilities demonstrate the power of deep context in enterprise AI workflows.
Whats next for SentinelForge
- Expanding automated pull-request reconciliation across Bitbucket.
- Introducing multi-tenant compliance reporting dashboards directly in Confluence.
Built With
- atlassian-forge
- atlassian-rovo
- confluence
- javascript
- jira
- model-context-protocol
- node.js
- teamwork-graph
