Ticket Intelligence

Connect your GitHub Issues, Jira, or Azure DevOps (Linear coming soon) and Mapnostics adds AI-powered effort estimates, impacted file predictions, and similar past fixes — directly on each ticket, in real time.

How it works

When you open a ticket in Mapnostics's Ticket Intelligence view, the AI reads the ticket title and body, searches your repository's codebase for related files and functions, and cross-references your git history for similar past issues. All without any manual tagging or labeling.

Related files — the top three files the AI believes the fix touches, verified against the code it was actually shown: a suggested file it never saw is dropped rather than shown as a hallucinated result.
Effort estimate — hours read from the ticket and the affected code, with a 15% buffer folded in and a complexity rating.
Suggested approach — a short plan for the fix, generated once the related files are known.
Impact warning — the true blast radius of the related files, read from the stored dependency graph rather than guessed.
Similar past fixes — tickets from git history that touched the same files.
Attachment context — text a vision model reads out of any screenshots attached to the ticket.

Live demo

Watch Ticket Intelligence analyse a real bug report — from idle to full AI analysis with effort estimate, affected files, and past similar fixes. Window height stays fixed throughout.

mapnostics.com/repo/mapnostics-app/tickets

Ticket Intelligence

AI-powered estimates, file impact, and similar past fixes

● LIVE

GitHub Issues✓ JiraConnect Azure DevOps

bugOpen#7

Dead code detection showing unexpected error on analysis

Running dead code scan on a repo with >500 files causes a timeout. Repro: connect any large repo and click "Analyze". Expected: scan completes. Actual: 504 timeout after 30s.

AI is analysing the ticket…

Effort Estimate2.5h

FILES LIKELY AFFECTED · 0

src/services/dead-code.service.tsHIGH

src/controllers/analyze.controller.tsMEDIUM

src/lib/batch-processor.tsLOW

SIMILAR PAST FIXES

→ Zombie scan timeout on 300+ files — fixed in #42

→ Analysis hanging on monorepos — fixed in #38

Effort estimates

How estimates are calculated

The AI reads the ticket description, identifies the likely affected functions and files using the codebase map, and estimates effort based on:

File count and size of likely-affected code.
Complexity of the affected functions (cyclomatic complexity).
Historical effort from similar past tickets in the same files.
Blast radius — how many other files depend on the affected code.

Accuracy

Effort estimates improve over time as Mapnostics learns from your team's actual cycle times. Estimates are calibrated to a senior engineer working alone — adjust up for junior engineers or unfamiliar areas.

NOTE

Effort estimates are advisory only. Use them for sprint planning and sizing conversations — not as performance metrics or commitments.

Ticket sources

Three sources are connected from the same bar at the top of the tab, and switching between them is instant once each is connected:

GitHub Issues — the default source. Selectable with no setup beyond the GitHub connection the repository already uses.
Jira — connect with the Connect Jira button, which starts an OAuth flow. Once connected, Jira becomes a switchable source alongside GitHub.
Azure DevOps — connect with a personal access token in the Connect Azure DevOps dialog, then pick a project.

NOTE

Linear is not yet a connected source.

TIP

Even without a formal ticketing system, you can paste any text description into the Ticket Intelligence search bar and get an instant AI analysis — useful for Slack messages, emails, or informal bug reports.