skTrace

[0.1.4] - 2026-06-10

• Added an out-of-date warning on shared reports. Each report now records the skTrace version that produced it, and the website checks that version against the latest release on Modrinth - when a report came from an older build, it shows a banner at the top prompting an update, with a link to the current version. Reports made before this change carry no version, so the site falls back to recognizing an older build from the report's own data, and those are flagged too. A report from a current build (or a dev build ahead of the release) is never flagged.
• Fixed (properly this time) Skript events firing twice after a script reload while skTrace was profiling. 0.1.3 detected reloads by watching for the /sk reload command, which missed every reload that wasn't typed by a player or the console - a script running execute console command "sk reload ...", a panel plugin or addon reloading scripts programmatically, and similar. skTrace now registers with Skript's own script load/unload events (modern Skript), so Skript's real triggers are restored before any unload - no matter how the reload was started - and re-hooked once loading settles. On Skript versions without those events, the 0.1.3 command detection remains as a fallback. Thanks Jake!
• Fixed the failure mode that made the doubling permanent - surviving profiler stops and even plugin updates, and fixable only by a full server restart. When a reload slipped past 0.1.3's detection and the profiler later stopped (or skTrace was updated via a plugin manager, which stops it on unload), skTrace wrote its backed-up trigger entries back into Skript's registries by list position. The registries had changed since the reload, so originals from the unloaded script generation could land back inside - or on top of - the new generation's entries, as orphans Skript can never remove. Every restore is now identity-checked: an original goes back only into the exact spot that still holds its skTrace wrapper, and anywhere Skript has rebuilt is left untouched.
• Fixed the same trigger being wrapped twice when it was reachable through two of Skript's registries (the event map and the script structures). Both now share a single wrapper - which also means that even if a reload were somehow missed entirely, Skript's own unload cleanup now finds and removes skTrace's wrapper by itself, so duplicate firing can't take hold in the first place.