feat(ptd): authed-route PTD tooling — PTD render-GET can't exercise authed routes (302/login); authed pages + authed-only behavior (toast/error-shown producer, práctica gates) never runtime-verified. Fix: in-process self-check route (preferred) or least-priv stored session for manual authed pass. Sibling: MARS-118.

Sub-tickets

Questions

Activity

• wi-cli-whey created · 2026-06-15
• wi-cli-whey commented · 2026-06-15
  VENUS-10 two-sided verification note (toast/error-shown, authed-browser required): see VENUS-10 comment — invalid práctica submit → FieldErrors + one toast/error-shown row; valid create → zero toast rows. Baseline pre-traffic = 0. This WI is the tooling fix that would let PTD drive that check automatically post-go-live.
• wi-cli-venus note · 2w ago
  DESIGN CONSTRAINT from probe-contamination measurement (coder02-venus-cc-ms8fdfq3u55b, 2026-07-31): today venus's telemetry recorder is unreachable from probes ONLY via the session gate — template.tsx sits at APP ROOT (no (protected) group; venus's mechanism differs from pluto's, coincident conclusion), and logRouteUsage's if(user) is the load-bearing gate. THE MOMENT this WI gives a probe a session, that gate passes, /api/* exclusion stops applying (target is a page), and the UA filter at usage-telemetry.ts:105 becomes the SOLE remaining defence. So: any authed-PTD implementation MUST carry UA evolutiva-internal-probe/venus-ptd (probe-curl.sh v2.0 --app venus) as an acceptance criterion, or authed PTD will pollute usage telemetry as real pageviews from day one.
• wi-cli-venus note · 2w ago
  CONSTRAINT UPGRADED (coder02-venus-cc-ms8fh8ikvfkc + coder-pluto-cc-ms8fh9ca8iuc): the authed-untagged-probe contamination class is UNCORRECTABLE once landed — webusage.usageEvents stores no UA/IP/header column, so post-hoc separation from real pageviews is impossible, not hard. Therefore the UA discipline must hold BEFORE this WI ships, not alongside: it is a PRECONDITION, the only control that exists. Benign instance of the exact shape: venus /admin/health n=2 2026-07-12 (real admin pageview, structurally identical to a contaminating probe). OPEN DESIGN QUESTION FOR ELAZAR (do not build): recording a UA-derived is-probe flag at emit would create the only possible post-hoc discriminator, but it is a privacy/telemetry-scope decision (what the logger auto-stamps), not a coder's call. Asked via DM 2026-07-31.