Update the kilo enkrypt integration for new models

[intentionally-left-nil]

Summary

We are tracking more models (see https://anaconda.atlassian.net/browse/AISAGE-35 )
and want to have this data shown on the kilo dashboard as well.

To do so, we need to update the mapping of enkrypt models to the model id.

Verification

Will double check the cron job after it runs, but the existing mappings were pulled from both enkrypt as well as the kilo leaderboard
curl -sS --max-time 30 'https://api.enkryptai.com/leaderboard/v2/scores'

and
curl -sS --max-time 20 'https://api.kilo.ai/api/gateway/models'

[kilo-code-bot]

WARNING: laguna-s-2.1 is mapped with source: 'xAI' while its canonical target is poolside/laguna-s-2.1.

source is part of the exact identity key used by matchEnkryptScores (identity.source === score.source), and Laguna models are published by Poolside — the sibling entry laguna-xs-2.1 below correctly uses source: 'poolside'. If Enkrypt reports poolside for this record, the mapping never matches and the model's stats are dropped as unreviewed_identity. The fixture at apps/web/src/tests/fixtures/enkrypt-scores.ts:154 mirrors the same value, so the test suite cannot catch it.

Suggested change

	identity: { model_name: 'laguna-s-2.1', provider: 'openai_compatible', source: 'xAI' },
	identity: { model_name: 'laguna-s-2.1', provider: 'openai_compatible', source: 'poolside' },

---

Reply with @kilocode-bot fix it to have Kilo Code address this issue.

[intentionally-left-nil]

The feedback is not accurate. Fresh live API call (seconds ago) confirms Enkrypt reports:
laguna-s-2.1 | provider: openai_compatible | source: xAI
laguna-xs-2.1 | provider: openai_compatible | source: poolside
Our mapping's source: 'xAI' for laguna-s-2.1 is exactly what the real leaderboard returns — captured from the live response when I added the mappings, not assumed. Enkrypt's source field reflects where they served the model for evaluation, not the publisher: they ran Laguna S 2.1 through xAI's endpoint and Laguna XS 2.1 through Poolside's, hence the sibling entries differ.
If we "corrected" it to 'poolside' per the feedback, the identity (laguna-s-2.1, openai_compatible, poolside) would never match any real Enkrypt record — creating exactly the silent unreviewed_identity drop the reviewer's warning describes. The fixture mirroring the mapping is by design (it encodes reviewed identities), so the test suite enforcing mapping↔fixture parity is working as intended; live API validation like this is the actual safety check.
The canonical target poolside/laguna-s-2.1 is right — only the identity triple must match Enkrypt's record, and it does. No change made.

[kilo-code-bot]

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Files Reviewed (4 files)

• apps/web/src/app/api/cron/sync-model-stats/route.test.ts
• apps/web/src/lib/model-stats/enkrypt-identity.test.ts
• apps/web/src/lib/model-stats/enkrypt-identity.ts
• apps/web/src/tests/fixtures/enkrypt-scores.ts

Previous Review Summary (commit 7c07454)

Current summary above is authoritative. Previous snapshots are kept for context only.

Previous review (commit 7c07454)

Status: 1 Issue Found | Recommendation: Address before merge

Overview

Severity	Count
CRITICAL	0
WARNING	1
SUGGESTION	0

Issue Details (click to expand)

WARNING

File	Line	Issue
apps/web/src/lib/model-stats/enkrypt-identity.ts	392	laguna-s-2.1 identity uses source: 'xAI' but maps to poolside/laguna-s-2.1; sibling laguna-xs-2.1 uses source: 'poolside'

Files Reviewed (4 files)

• apps/web/src/lib/model-stats/enkrypt-identity.ts - 1 issue
• apps/web/src/tests/fixtures/enkrypt-scores.ts - mirrors the same mapping (fixture confirms tests cannot catch it)
• apps/web/src/lib/model-stats/enkrypt-identity.test.ts - 0 issues
• apps/web/src/app/api/cron/sync-model-stats/route.test.ts - 0 issues

Fix these issues in Kilo Cloud

---

_{Reviewed by deepseek-v4.1-flash · Input: 0 · Output: 0 · Cached: 0}

_{Review guidance: REVIEW.md from base branch main}