Ops Triage — the 37 'planned' (strategy) docs: done / parked / dead

2026-07-22 · Ben (audit worker) · report

Ops Triage — cutting the 37 "planned" docs down to what's actually live

The problem: type: strategy is a graveyard. 37 docs, only 7 with any status, 30 with none — so shipped plans, dead ideas, and live work all pile into one "planned" list and you can't find anything. (Digests are the only clean group because they're append-only + dated.)

The fix (below): classify all 37, archive the done + dead, split living references out of "planned," leaving ~8 genuinely-active plans. Verified the no-signal ones against the actual codebase so nothing live gets killed.


Verdict summary

VerdictCountWhat to do
DONE (shipped/verified)20status: done → move to archive/
DEAD (never built / superseded / not worth it)4decide: revive or delete
PARKED (real, deferred)4status: parked → keep, out of the active view
REFERENCE (living, not a "plan")3retype reference → stop cluttering "planned"
ACTIVE (current work)6keep in "planned" — this is the real list

Net: "planned" goes from 37 → ~8 (6 active + a couple parked you'd still glance at).


DONE → archive (20)

DocWhy done
forms-upgradestatus:done; shipped
responsive-ctastatus:done; shipped
wcag-accessibility-briefdone + promoted (site is 0-violations, gated)
content-differentiationdone + superseded by later content work
engagement-toast-moduledone; shipped
domain-broker-action-plan-validationdone; shipped; stale
conversion-tracking-architecturedone; obsolete (server-side conv. live, note #177)
provider-plan-streamlinecomplete; shipped
domain-broker-recoverycomplete; shipped
domain-broker-funnel-restructurecomplete; stale
domain-broker-final-roundcomplete build spec; broker shipped
domain-broker-hero-ab-launchsuperseded by domain-broker-v2-launch-runbook + note #180 (v2 live)
astro-7-tailwind-4-migrationpromoted to prod (note #173)
advisor-instrumentationshipped (note #179); advisor UI is a separate concept
reviving-hosting-clusterexperiment concluded (hosting revived)
next-50-daysdated 2026-06-04 — timeframe expired
meeting-takeaways-2026-07-07past meeting; superseded by Jul 21 takeaways
critical-vpn-roi-gateway-templategateway shipped (vpn-out/src/gateway.js); tracked in memory
hero-trustscore-redesignTrustScore hero shipped (note #177). Follow-up only: the Jul 21 "open score breakdown by default + swap assessment/pricing" tweak — split that into a fresh 1-line active item; archive the rest
final-shipment-julyJuly plan, mostly executed — confirm the last items landed, then archive

DEAD → decide (revive or delete) (4)

DocReality check
vpn-api-docs-hubNever built — no src/pages/api*/docs* hub exists (only internal src/data/API.md). Untouched since 2026-05-27. Kill unless the public API hub is still wanted.
perks-points-rewardsNever built — no points/loyalty code anywhere. Untouched since 2026-06-09. The "anti-churn loyalty engine" never happened. Kill or revive as a real initiative.
comparison-superclusterStalled/partial — a handful of vs-pages exist (proxy-vs-vpn, wireguard-vs-openvpn, com-vs-xyz) but not the provider-vs-provider matrix the doc plans. Decide: finish the matrix or drop the ambition.
google-ads-conversion-value-strategySuperseded by google-ads-value-loop-plan (its own frontmatter says superseded). Archive.

PARKED → mark + keep (4)

DocNote
text-to-speechClient-requested, deferred. Real, not now.
accessibility-toolbarClient-requested; partially shipped, rest parked.
google-ads-value-loop-planIn-progress — $1 lead leg fixed, qualified→5% leg still unvalidated (memory ads-offline-ghl-watch). Keep as active-watch, not "planned."
speed-labStill relevant + still NOT done — the pipeline emits value: 'Not tested'; Speed Lab data is fabricated. This is the root of the FAQ provenance liability (43 Mbps + 14 "we tested" answers). Reclassify as active if we act on provenance; else parked.

REFERENCE (living strategy, not a task) → retype reference (3)

DocWhy
geo-ai-citationsStanding GEO/AI-citation strategy; referenced by the FAQ work. Not a plan to "finish."
authority-roadmapStanding "Reviewed by VPN.com" authority strategy.
(speed-lab)If kept as strategy-reference rather than an active build — author's call.

ACTIVE — the real "planned" list (6)

The FAQ overhaul + the conversion principle, all current:

Plus, once split out: the provider-template tweak (from hero-trustscore-redesign) and speed-lab/provenance if we green-light it.


Reorg recommendation (so this doesn't rot again)

  1. Enforce status on every strategy doc: active | parked | done | dead. This one field is the whole fix — the editor can then default the "planned" view to status: active.
  2. done + deadarchive/ (already exists with 7). Keeps history, clears the list.
  3. Split type: reference for living strategy (geo-ai-citations, authority-roadmap) so standing docs stop masquerading as open tasks.
  4. One index front-door: an ops/README.md (or pinned index) listing only active plans + a link to archive — since digests are the only group you navigate by, give "planned" the same at-a-glance entry.
  5. Going forward: a plan doc gets status: done + moves to archive in the same commit that ships it (like the domain-broker runbook got its launch banner). The digests already work because they're disciplined; apply the same discipline here.

Bottom line: ~20 of the 37 are already shipped, 4 are dead, 3 are standing references. Only ~8 are live work. The list feels unfindable because 29 of 37 shouldn't be in the "planned" view at all.