POTESTAS 6.2 — the analysis face: recurrence, recognition latency, coverage

[4444J99]

Phase: #2084 · Program: #2077

Reuse scripts/studium-analysis.py — the existing "packaging as data analysis" face — rather than building a POTESTAS-specific dashboard.

What it answers

• Which mechanisms recur across observations — the personal threat profile, derived rather than assumed.
• Which counters hold — where a recorded counter was applied and the situation changed.
• Recognition latency — is a mechanism being named sooner over time? This is the program's actual training signal. Not "did I avoid it" but "how fast did I see it," which is the skill inoculation builds.
• Coverage — which registry families have never been observed. Absence is information: it means either a genuine non-exposure or a blind spot, and the two are worth distinguishing.

Fails open — a hard requirement

The existing analysis face renders corpus-level analysis from the authored curriculum even at zero events. Preserve that exactly. A face that shows nothing until data accumulates is a face nobody returns to, and this program has a long cold start by design (slow cadence, self-scored, no automation).

What it must NOT do

• No per-person aggregation. Events carry no identity, so this is enforced by the schema rather than by discipline — which is the point.
• No score, index, or trend line about any individual.
• No streak or engagement mechanic. Building a variable-reward loop into a curriculum about variable-reward loops would be the funniest possible failure and also a real one.

Done

studium-analysis.py renders the POTESTAS view, exits 0 with zero events, shows recurrence and recognition latency once events exist, and aggregates nothing per-person.