Run the gauges on your repo — help answer the population question (nulls welcome)

[dcnconsult]

This project has an unusual ask: don't star the repo — run the instruments.

SentAInce grew up gauge-first: every organ ships with the measurement that could kill it, and several shipped off because their own gauge said the prize was modest. The biggest open question isn't a feature — it's how these memory dynamics behave across many people's real work. That's a population question, and a single maintainer's repos cannot answer it. You can.

If you've installed the organism (quickstart) and worked with it for a while, you have an accrued corpus. The gauges are read-only and run in seconds:

python -m exocortex.deploy status /path/to/your/project
# is credit going to real work or routing noise?
python -m exocortex.gauge.credit_hygiene_gauge --state-dir /path/to/your/project/.claude/exocortex
# do abstain/veto events even occur in your traffic?
python -m exocortex.gauge.uncertainty_gauge --audit /path/to/your/project/.claude/exocortex/audit.jsonl

Post what you see — including nulls and boring results, which are as valuable as wins here. Interesting things to include: how many goal-classes your work minted, whether routes converged, your credit rate, anything that looks wrong.

One privacy note, honestly: gauge output is statistics, but a couple of fields can echo command text from your own audit trail — skim before posting, redact freely. Nothing leaves your machine unless you paste it.

What we'll do with the results: they feed the public evidence ledger's population columns — with credit, and with the same qualifiers we apply to our own numbers.