coulomb/fin-hub
4
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

feat: publish unblended AI-plan effectiveness series

Browse source 
Report booked cost, entitlement, token coverage, and work as three
series: work per measured token, work per measured+estimated token,
and work per euro. Months below 50% coverage are unfit for token
trends. There is no blended efficiency number.
This commit is contained in:
tegwick 2026-08-15 19:53:28 +02:00
parent ee90ba2873
commit 033ad3f41f
12 changed files with 512 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

7
docs/ai-plan-token-effectiveness.md
View file

@ -106,8 +106,11 @@ State Hub `measurement_kind` maps as:
| no events | `unknown` residual, not a zero observation |
A derived overlay (implied €/token, work per token) is only as strong
as its weakest input. A month whose coverage is incomplete is
`insufficient` for trend comparison, not interpolated.
as its weakest input. Coverage is measured / (measured + estimated).
A month whose coverage is below `0.50` is `unfit` for token-series
trend comparison, not interpolated. The effectiveness report publishes
three series — work per measured token, work per measured+estimated
token, and work per euro — and never a blended efficiency number.
## What fin-hub will not do