Grok Build Update: Enhanced Long-Running Agent Workflows and Enterprise Controls
XFreeze · x · 2026-09-01
Grok Build released versions v1.0.16 and v1.0.15, significantly improving resilience for long-running agent workflows and overall speed.
Key Updates:
- Performance Boost: Model connections, token refreshes, and startup tasks now run in the background, reducing first-reply latency and speeding up new sessions and exits.
- Long-Running Tasks: Supports workspace-level MCP integrations; improves recovery from token/network interruptions to prevent data loss; ensures more reliable subagent spawning.
- Enterprise Controls: Enterprise policies can restrict model selection via requirements.toml; fixed MCP OAuth flows.
- UX Improvements: Fixed blank labels for long command suggestions; optimized memory consolidation during session exit.
More from coding & agent
- Coinbase Launches x402 Payment Protocol for API-Free AI Agent Data Access — kleffew94 · 2026-09-01
- Dev finds asking AI to grill you first improves results, even if you'll accept everything — lucasmeijer · 2026-09-01
- Delegating to Agents Requires a Leap of Faith Despite High Stakes — manosaie · 2026-09-01
- RPM-guided search sets new SOTA on benchmarks, proving selection scales with compute — anirudhg9119 · 2026-09-01
- Git vs vector DB for MCP agent memory: where each breaks in production — Future_AGI · 2026-09-01
- Google's Antigravity Teamwork + Gemini 3.7 Flash Solves 7 Open Math Problems — rseroter · 2026-09-01