Lawful AI Daily Brief — 2026-04-18

🛠️ Tool Updates

• Claude Code v2.1.114 fixed a permission-dialog crash in agent-teams flows.
• Claude Code v2.1.113 pushed a chunky hardening pack:
  • native per-platform binary launch
  • sandbox.network.deniedDomains for explicit network blocking
  • stricter Bash policy matching (sudo wrappers, safer find -exec/-delete, stricter risky rm paths on macOS)
  • MCP concurrent-call timeout fix and smoother /loop + /ultrareview behavior
• Codex tightened trust boundaries in .codex for hooks/exec policies and improved plugin controls + skill-context budgeting.

💡 Tip of the Day

Treat trust as code, not vibes.

# .codex/config.toml
[project]
trusted = false

[security]
requireTrustForHooks = true
requireTrustForExecPolicy = true

Why this slaps: it blocks “surprise automation” until the project is explicitly trusted. Safer defaults, calmer mornings ☕

🔥 Trending Repos

Top new repos (recent):

• zats/permiso (Swift) — permission dialog UX for accessibility settings
• Fleet-Management-System (Java) — newly active project
• SKILL-LINK-AI (TypeScript)
• OHLC-Vol (Python) — volatility forecasting from OHLC estimators

AI/LLM movers:

• Significant-Gravitas/AutoGPT
• langgenius/dify
• open-webui/open-webui
• firecrawl/firecrawl
• infiniflow/ragflow
• bytedance/deer-flow
• mem0ai/mem0

🎤 Standup One-Liner

"We hardened agent tooling defaults, tightened trust boundaries, and kept our stack fast and less spooky in production." 👻✅

---

Repo: lawful-ai-staging