Lawful AI Daily Brief — 2026-04-30

⚙️ Tool Updates

• Claude Code v2.1.123 fixed the OAuth 401 retry loop when experimental betas are disabled. Less auth drama, more shipping.
• Claude Code v2.1.122 added ANTHROPIC_BEDROCK_SERVICE_TIER, smarter /resume from pasted PR URLs, and cleaner /mcp behavior.
• Codex CLI dropped 0.126.0-alpha.15 → .17 quickly, with packaging that now looks like a mini AI runtime stack (codex-app-server, command runner, responses proxy).

💡 Tip of the Day

If you want clean audit breadcrumbs for compliance reviews, hook every tool call into a tiny JSON log:

# audit_log.py
import json,sys,time,hashlib
e=json.load(sys.stdin)
row={
 "ts":int(time.time()),
 "tool":e.get("tool_name"),
 "prompt_hash":hashlib.sha256((e.get("input","")+"").encode()).hexdigest()[:16],
 "output_id":e.get("tool_use_id")
}
print(json.dumps(row))

⚖️ Legal x AI Watch

• artvana/global-ai-atlas was updated — structured, machine-readable global AI regulation tracking.
• kody32/eu-ai-act-guide was updated — practical EU AI Act + GDPR guidance for startups/SMEs.
• Alvoradozerouno/GENESIS-v10.1 was updated — EU banking compliance framing with AI Act links.
• simaba/ai-prism was updated — curated responsible-AI governance resources.

📚 Fresh Papers

• Turning the TIDE: Cross-Architecture Distillation for Diffusion Large Language Models
  https://arxiv.org/abs/2604.26951
• ClassEval-Pro: A Cross-Domain Benchmark for Class-Level Code Generation
  https://arxiv.org/abs/2604.26923
• ClawGym: A Scalable Framework for Building Effective Claw Agents
  https://arxiv.org/abs/2604.26904

🔥 Trending Repos

• f/prompts.chat — ⭐ 161k — still the prompt bazaar boss.
• langgenius/dify — ⭐ 139k — agentic workflow platform keeps climbing.
• langchain-ai/langchain — ⭐ 135k — no signs of slowing down.
• NousResearch/hermes-agent — ⭐ 123k — agent framework momentum stays strong.
• bytedance/deer-flow — ⭐ 64k — long-horizon super-agent harness keeps trending.

🎤 Standup One-Liner

“Today we tightened our Claude/Codex stack and turned compliance logging from ‘nice idea’ into copy-paste reality.”

---

Source repo: https://github.com/laugustyniak/lawful-ai-staging