About
What Vigil is.
Vigil is a dashboard that ingests FFLogs combat data for one high-end FFXIV static and turns it into per-pull, per-mechanic signal during progression on an ultimate.
It exists to answer three questions:
- Where are we losing pulls, and whose originating mistake caused each wipe? Not chain reactions — root causes. Vigil distinguishes roots from cascades and from mit failures, with strat-aware refinement when you author a plan.
- Are we DPS-gated or mechanics-gated?The answer is per phase. Vigil derives an empirical DPS check from clearing groups' phase-by-phase raid DPS and compares it to yours.
- How does our prog compare to the field, on the parts that are actually comparable? Boss-side outcomes only — never positioning or strat decisions, since those are group-specific.
Vigil also builds a crowd-sourced model of each encounter's deterministic boss-side structure (cast order, phase timing, mechanic types) since no official timeline exists at launch. The model gets refined automatically as new reports come in.
Who it's for
Vigil is built for a high-end static progging an ultimate. The primary user is a raid lead who reads the dashboard after each session, identifies what's killing the group, and adjusts the next session's focus accordingly.
It assumes you already understand FFLogs — what a parse is, what aDPS means, what a tankbuster vs. raidwide looks like. It does notassume you understand the encounter yet; that's what the fight-model inference is for.
Why it exists
Off-the-shelf FFLogs tells you what happened. It doesn't tell you what to fix. The percentile is for clears; during prog there is no parse. Most groups make decisions from gut feel about which wipe was "whose fault" and which mechanic is "the wall this week."
Vigil exists to replace gut feel with measurement. Every metric it surfaces is derived from raw event data and traceable back to a specific cast or death. When it says "Cyclonic Break killed your party in 6 of 12 occurrences", that's not an opinion.
Status
Vigil is post-1.0.0 and actively developed. The full Phase 0–3 feature set has shipped: ingestion, fight-model inference, fault attribution, mit audit, DPS-gate diagnostic, optimization mode, session reports, and a React dashboard. Recent work is on the consumer side — multi-static support, cleaner onboarding, and per-static FFLogs OAuth.
See the news page for the running changelog, or read the docs to set it up locally.