July 2026: full registry refresh, 40 new evaluations, Guard0 design system, State of AI Trust report - #29
Merged
Merged
Conversation
Every evaluation re-verified against primary sources on 2026-07-09: lifecycle statuses (retirements, deprecations, supersessions), pricing, versions, CVEs, and hosted-endpoint changes. Fixes dangling cross-references in 15 files (claude-4-opus, gpt-4-5, etc. mapped to real registry ids or dropped). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Models: Claude Sonnet 5, GPT-5.6 (Sol/Terra/Luna), Grok 4.5, GLM-5.2, MiniMax M3, Kimi K2.7-Code, Nemotron 3 Ultra, Qwen3.6. Agents: Claude Cowork, OpenClaw, ChatGPT Agent, Microsoft Scout, Perplexity Comet, Poke, Trae, Cline, OpenCode, Goose, Warp, Junie, Factory Droids, Replit Agent, Lovable, Antigravity, Kiro. MCPs: Snowflake, Databricks, Salesforce, HubSpot, Asana, Box, Shopify, PayPal, Canva, Grafana, Neon, Exa, Browserbase, ClickHouse, Neo4j. Registry grows 156 -> 196. All scores evidence-linked and calibrated against existing siblings. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace the hand-maintained 158-import lib/data.ts with codegen: scripts/generate-data-index.ts scans data/ and emits lib/data-index.ts (full entities, server-side) and lib/data-summaries.ts (lightweight EntitySummary literals for client views). List/compare pages now ship ~158KB of summaries instead of ~4.4MB of full JSON: homepage first-load JS drops 665kB -> 138kB (-79%). Also: draft gate (_-prefixed files or "draft": true stay unpublished), identifier-collision detection, predev/prebuild/pretype-check/prevalidate hooks, CI drift check for the generated index, memoized EntityCard, and cached overall scores for sorting. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Emerald #10B981 primary on FAFAFA/true-black surfaces, Plus Jakarta Sans
display + Inter body + JetBrains Mono (variable fonts vendored locally
for offline builds), guard0 dog-mark logo lockup ('TrustVector by
Guard0'), sentence-case hero matching guard0.ai, dot-grid, small radii,
elevation + motion system with prefers-reduced-motion support.
Scores render as a semantic traffic-light ramp (emerald/light-emerald/
amber/orange/red) from a single SCORE_THEMES source of truth; badges,
bars, and segments all derive from it. Fixes: UTC date rendering
(July 8 off-by-one), animation fill-mode pinning the card hover lift,
duplicate box-shadow overriding the hover glow, search-icon overlap,
stale sky-blue Strong leftovers, dead Loot Drop-era CSS. README badges,
findings, and screenshots updated for the 196-entity registry.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Standalone industry briefing generated from the 196-evaluation registry: model lifecycle churn, the capability-trust gap across stack layers, the 2026 agent security crisis, MCP professionalization and its failure modes, vibe-coding security debt, and ownership churn — with scoreboards, watchlist, and a practitioner playbook. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
✅ Deploy Preview for scintillating-liger-b66f71 ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
Replace the old shield favicon with Guard0's dog favicon set (favicon.ico / 32x32 / 192x192). Convert the contribute, methodology, and compare page heroes to the guard0.ai treatment (sentence-case Plus Jakarta Sans extrabold with emerald keyword) and remove the uppercase transform from 41 headings site-wide so the display face reads as it does on the main site. Micro-labels and eyebrows stay uppercase per the guard0 label voice. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
… status ticker
The registry becomes the interface:
- Registry Field: hero renders one dot per evaluated system (real data),
colored by trust tier and sorted into a gradient of trust; hover names
the system, click opens its evaluation
- Trust Spectrum: every entity card carries a five-band fingerprint (one
band per dimension, tier-colored) along its top edge, growing on hover
- Registry Feed: a mono security-ticker strip of real lifecycle events
(retirements, deprecations, sub-50 trust floors) derived from the data,
pausing on hover, linking to each record
- Dossier stamps on detail pages ('Evaluation record · <id>') and a faint
emerald bloom behind the hero headline
All motion honors prefers-reduced-motion; ticker classification is
prefix-strict so supersessions never read as retirements. README header
screenshot refreshed.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Five workstreams in one branch (one commit each):
1 ·
data:July 2026 registry refreshAll 156 existing evaluations re-verified against primary sources on 2026-07-09 — lifecycle statuses, pricing, versions, CVEs, hosted endpoints. Highlights: Claude Fable 5's 19-day export-control suspension scored in; Gemini 3 Pro marked retired; GPT-4.1 family context corrected (128K → 1,047,576); Grok 3 silent-redirect pricing documented; Langflow/Flowise/n8n exploited-CVE waves recorded; 15 files' dangling cross-references fixed.
2 ·
feat:40 new evaluations — registry grows 156 → 1963 ·
feat:auto-generated data index + client summary bundlescripts/generate-data-index.tsreplaces the hand-maintained 158-import list: drop a JSON intodata/and it publishes (with a_-prefix/"draft": truegate for WIP files). Client views now ship a lightweight generated summary instead of full evaluations: homepage first-load JS 665kB → 138kB (−79%). Includes identifier-collision detection, npm pre-hooks so fresh clones/CI work, and a CI drift check for the generated index.4 ·
feat:Guard0 design system migrationEmerald
#10B981on#FAFAFA/true-black, Plus Jakarta Sans display + Inter + JetBrains Mono (variable fonts vendored — offline builds), dog-mark "TrustVector by Guard0" lockup, guard0.ai-matched hero, semantic traffic-light score ramp from a singleSCORE_THEMESsource of truth, elevation/motion system with reduced-motion support. Fixes found in review: UTC date off-by-one, animation fill-mode killing card hover, dead hover glow, search-icon overlap, stale sky-blue ramp leftovers, dead Loot-Drop CSS.5 ·
docs:State of AI Trust — July 2026Standalone intelligence briefing generated from the registry (
docs/reports/STATE-OF-AI-TRUST-2026-07.md): model die-off, the capability-trust gap by stack layer, the agent security crisis, MCP failure modes, vibe-coding security debt, ownership churn, watchlist + practitioner playbook.Test plan
npm run validate— 196/196 evaluations validnpx tsc --noEmit— cleannpm run build— 204/204 static pages🤖 Generated with Claude Code