fix: add null checks for tactic/technique toLowerCase calls

[impsislegobatman]

Fixes #200

Added optional chaining (?.) before .toLowerCase() calls on tactic and technique fields that the backend sometimes returns as null.

[Polliog]

Hi @impsislegobatman, thanks for taking a stab at this!

A couple of notes before we can merge:

1. The crash reported in [BUG] Cannot read properties of null (reading 'toLowerCase') #200 originates in packages/frontend/src/lib/components/siem/dashboard/MitreHeatmap.svelte:40, where tactic.toLowerCase() is called directly on a cell's tactic field before getTacticName() is ever invoked. So adding ?. inside the shared getTacticName / getTacticDescription / etc. won't actually prevent the reported crash: it would still throw one line earlier in abbreviateTactic.

2. The change in packages/backend/src/modules/sigma/routes.ts guards params.tactic?.toLowerCase(), but params.tactic is a required route path parameter, so if the route matches it's always a string and the optional chaining there is unreachable.

3. We've confirmed with the reporter (see [BUG] Cannot read properties of null (reading 'toLowerCase') #200) that the null values come from actual detection events stored in the DB after importing the "Auth and Security Pack" community rules, so the real fix needs to either drop those cells in the heatmap or backfill a sensible default.

We've already pushed a fix in release/0.9.2 that filters out cells with null tactic/technique in MitreHeatmap and hardens DetectionEventsList.getLogLevelClass. That said, your defensive ?. additions in shared/utils/mitre.ts are reasonable hygiene for other callers. If you want, we can keep that part and drop the other two changes. Let us know!

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[Polliog]

Thanks for following up on the feedback so quickly! The trim to keep just the defensive changes in shared/utils/mitre.ts is exactly what we discussed clean second commit too.

Quick note on context: the actual crash from #200 (Cannot read properties of null from the MITRE heatmap) was fixed in #203 by filtering out null tactic/technique cells in MitreHeatmap.svelte, which just landed on main. So #200 is already resolved at the source. Merging this PR as additional hardening your ?. additions still help any future caller of these helpers that might receive a null/undefined value, so it's good defensive code to have.
Appreciate the contribution! 🙌