PM-39866 - Enhance seeder to seed archive and deleted ciphers

[theMickster]

🎟️ Tracking

PM-39866

📔 Objective

Adds archive and delete lifecycle-state generation to the Seeder's density-driven cipher pipeline for both personal and organizational vaults. New DensityProfile rate + ceiling fields drive the feature similar to the pre-existing properties. Altered the LoginCipherScene.cs to handle archived ciphers.

🧪 Testing

Expand for details

End-to-end via the SeederApi scenes

Drove the SeederApi /seed endpoint end-to-end to build a full account graph and confirm the new
archive/delete request flags flow through the scene → factory → encrypted-entity path:

• SingleUserScene + SingleOrganizationScene / OrganizationUserScene — create the
  user, the org, and org membership.
• UserLoginCipherScene and OrganizationLoginCipherScene with Archived: true and
  Deleted: true — create archived and deleted login ciphers in both the personal and org
  vaults. Archived writes the per-user Archives JSON via CipherComposer.BuildArchivesJson
  (same shape as Cipher_Archive.sql); Deleted stamps DeletedDate.
• Various cipher types via UserCardCipherScene, UserIdentityCipherScene,
  UserSecureNoteCipherScene, UserBankAccountCipherScene, UserDriversLicenseCipherScene,
  UserPassportCipherScene, UserSshKeyCipherScene — confirm the mixed-type creation path still
  seeds and encrypts correctly alongside the archived/deleted login ciphers.

Confirmed the seeded archived/deleted login ciphers surface correctly in the web-vault UI (Archive
and Trash tabs) when logged in as the seeded user and org owner.

Live preset seed + SQL verification

Seeded Scale presets via dotnet run -- preset --name scale.<name> --mangle and queried
[dbo].[Cipher] directly (verification.md Q9 + Q11) for each. Every count landed exactly on the
configured target/cap — no overshoot, no starvation:

Preset	Org ciphers	Deleted (Q9)	Archived (Q11)	Both	Notes
Central Perk (XS)	200	4	8	2	percentage targets (2% / 4% / 1%)
Bluth Company (SM)	500	20	40	15	percentage targets (4% / 8% / 3%)
Umbrella Corp (MD)	3,000	25	50	25	clamped to 25 / 50 caps; all
Tyrell Corp (LG)	17,000	25	50	25	clamped to 25 / 50 caps; all deleted also archived

Step 4: Owner-visibility guarantee (live)

Logged into the web-vault UI as the seeded org Owner and confirmed the Owner always has at least one
collection-assigned archived item and one deleted item visible — the Archive and Trash tabs
render them, matching the SQL counts. Verified for Central Perk and Bluth Company.

[github-actions]

🤖 Bitwarden Claude Code Review

Overall Assessment: APPROVE

Reviewed the Seeder density-pipeline enhancement that adds archived/deleted cipher lifecycle-state generation for both personal and org vaults. Traced the target-math (ComputeTargets), index selection (Select/EvenlySpacedIndices), round-robin owner attribution, and the deferred resolveArchiveOwner callback path in both GenerateCiphersStep and GeneratePersonalCiphersStep. This is dev/test tooling under util/Seeder with no production runtime or vault-crypto impact.

Code Review Details

No blocking findings.

Notable strengths that informed the assessment (no action required):

• canArchive=false correctly zeroes both archivedTarget and bothTarget, so the deferred resolveArchiveOwner callback is never invoked with an empty position map.
• Replacing EstimateTotal with the exact runningOffset makes expectedTotal precise rather than estimated, improving typeDistribution.Select bucketing.
• Archived/deleted DaysAgo windows (15–90 vs 7–14) never overlap, guaranteeing ArchivedDate < DeletedDate for "both" ciphers, and this is locked in by regression tests.
• New logic is backed by focused unit and regression tests in ArchiveAndDeleteRateTests and CipherComposerTests.

The three prior github-code-quality threads (unused locals, constant condition) are resolved in b9af380.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 61.36%. Comparing base (e899a9c) to head (b9af380).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #7923   +/-   ##
=======================================
  Coverage   61.35%   61.36%           
=======================================
  Files        2236     2236           
  Lines       98547    98547           
  Branches     8911     8911           
=======================================
+ Hits        60468    60469    +1     
+ Misses      35943    35942    -1     
  Partials     2136     2136           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.