Your tracker.
Your repository. Your keys.

Syncs the epic→story→task hierarchy into a board, resolves parents in two passes, and writes new items and test cases back when you ask it to.

The same sync against Azure Boards, with Feature mapped onto epic so the hierarchy lines up with Jira projects.

Branch, commit and pull request. Stack detection reads the repository on connect, and the test-script commit stacks onto the existing PR rather than overwriting it.

Commits and pull requests against an existing file tree, for teams whose code lives beside their boards.

The first of two generation agents that runs on every Blue Mantis pipeline. Raptia is optimised for precision — it reads the work item, the acceptance criteria, and the repository context, then commits to a single well-reasoned answer. Stack-aware: detects React, Angular, Vue, Node.js, .NET, Java Spring Boot, Python, and Go — and writes idiomatic code for each without being told.

The second generation agent. Fovea takes a wider view of the same context — it considers more of the repository before settling on an approach, which means it often catches what Raptia misses. Also stack-aware — uses the same detected profile to ensure both suggestions follow the same framework conventions.

The ranking agent. After Raptia and Fovea complete, Synthesia scores both suggestions on correctness, readability, diff size, convention adherence, and acceptance-criteria coverage — then recommends the better one with a confidence score. Two behaviour signals — ambiguity handling and surgical precision — flag whether an agent silently assumed something or changed more than the work item required.

The review agent. After you commit a suggestion, Veria reads the committed code against the work item's acceptance criteria and produces a structured review: strengths, gaps, risks, and a one-sentence focus note for the human reviewer.

The security agent. Runs after commit using a frontier security model. Scans the committed change for OWASP Top 10 vulnerabilities, hardcoded secrets, injection flaws, and authentication bypasses. Outputs structured findings with severity, OWASP category, line reference, and remediation steps. High and Critical findings block the PR via a GitHub status check. Remediate Now creates a tracker ticket, syncs the board, and starts a new run to fix the issue — without leaving Blue Mantis.

The documentation agent. After a run completes, Narratia generates an operational runbook: a summary of what changed, deployment steps specific to this change, rollback procedure, validation commands, test cases from the generated suite, and security findings from Aegis. Pushed to Confluence, Notion, or committed as Markdown to the PR branch.