@rauchg: Harness durability in http://fx.sh. The same ๐šŒ๐š›๐šŽ๐šŠ๐š๐šŽ๐™ต๐šก๐™ฐ๐š๐šŽ๐š—๐š() is now *durable* on your laptop or in the cloโ€ฆ

X AI KOLs Following Tools

Summary

fx/libfx v0.0.14 introduces a durability adapter API that makes its Zig-based coding agent harness survive crashes, redeploys, and timeouts across local, Postgres, or Vercel Workflows backends with minimal integration code.

Harness durability in http://fx.sh. The same ๐šŒ๐š›๐šŽ๐šŠ๐š๐šŽ๐™ต๐šก๐™ฐ๐š๐šŽ๐š—๐š() is now *durable* on your laptop or in the cloud, via files, Postgres, or Vercel Workflows. Node.js or Bun. Agents need long-running compute. But they also need to tolerate restarts, redeploys, timeouts, crashesโ€ฆ Concretely, if a user sends a prompt and then throws their phone into a black hole, we ensure that turn runs to completion. If the process crashes mid-tool-call, it resumes from there. To this end, ๐š•๐š’๐š‹๐š๐šก introduces a simple durability adapter API. On Vercel, it leverages all the infra we built for durable workflows and queues, giving you reliability with 1LOC. This is a critical aspect of production-grade agents (and something all @eve agents also get out of the box). Next, we'll show you how we bring the same durability to Pi, OpenCode, and even your own custom harnesses.
Original Article
View Cached Full Text

Cached at: 10/17/26, 04:47 AM

Harness durability in http://fx.sh.

The same ๐šŒ๐š›๐šŽ๐šŠ๐š๐šŽ๐™ต๐šก๐™ฐ๐š๐šŽ๐š—๐š() is now durable on your laptop or in the cloud, via files, Postgres, or Vercel Workflows. Node.js or Bun.

Agents need long-running compute. But they also need to tolerate restarts, redeploys, timeouts, crashesโ€ฆ

Concretely, if a user sends a prompt and then throws their phone into a black hole, we ensure that turn runs to completion. If the process crashes mid-tool-call, it resumes from there.

To this end, ๐š•๐š’๐š‹๐š๐šก introduces a simple durability adapter API. On Vercel, it leverages all the infra we built for durable workflows and queues, giving you reliability with 1LOC.

This is a critical aspect of production-grade agents (and something all @eve agents also get out of the box). Next, weโ€™ll show you how we bring the same durability to Pi, OpenCode, and even your own custom harnesses.


fx - Tiny, open, native coding agent

Source: https://fx.sh/ v0.0.14(notes)6.66mibstatus:experimentaluse at your own risk, we will be making frequent changes

This demo runs the full fx CLI as WebAssembly compiled with the Zig toolchain. Networking is delegated to browserfetch, and every SDK aspect is configurable in thedocs.

fx is a coding agent harness and CLI written in Zig, optimized for research and embeddability as part of larger systems.

It focuses on minimalism and performance across the board, from system prompt design to its tools, feature set, and 6.66 MiB binary.

Its output is closer to a Unix shell than an IDE in the terminal.

It is open source (Apache-2.0), model-agnostic, and suitable for both local and cloud inference.

Tiny native binary

Install one binary for your terminal, CI, or agent sandbox.

Fast startup

Start an interactive session or run a single request withfx ask. No background service is required.

WebAssembly support

Run the agent or terminal in the browser withlibfx. Your application controls storage, tools, and network access.

Shell-like UI and ergonomics

fx preserves scroll history and keeps tool activity in the conversation. Open the full transcript when you need more detail.

Context efficient

Concise tool schemas keep the context small. Skill instructions and MCP tool schemas load when needed.

Embeddable and extensible

Embed the core withlibfx, add your own tools, and connect skills or MCP servers.

Pranit (@fazxes): fx v0.0.14

libfx is now durable out of the box. Same API on local, Vercel and DIY.

  • Sessions survive crashes, redeploys & timeouts
  • Power cloud agents with ChatGPT & Grok subscriptions
  • Custom providers & local model support in libfx
  • Run multiple libfx agents in one app

Similar Articles

Making every harness durable (7 minute read)

TLDR AI

Vercel ๅ‘ๅธƒๅผ€ๆบๅทฅๅ…ท durable-harness๏ผŒ่ฎฉ Piใ€OpenCodeใ€Codexใ€Claude Code ็ญ‰็ผ–็ ไปฃ็†ๅœจไบ‘็ซฏๅ…ทๅค‡ๆŒไน…ๆ€ง๏ผŒๅฏๅœจๅดฉๆบƒๆˆ–ๅ‡ฝๆ•ฐ่ถ…ๆ—ถๅŽไปŽๆ—ฅๅฟ—ๆขๅคไผš่ฏไธŽๅทฅไฝœใ€‚

HarnessX: A Composable, Adaptive, and Evolvable Agent Harness Foundry

Hugging Face Daily Papers

HarnessX is a foundry for composable, adaptive, and evolvable AI agent harnesses that uses compositional primitives and trace-driven evolution to improve agent performance. Across five benchmarks, it achieves an average gain of +14.5% (up to +44.0%), demonstrating that runtime interface evolution is a complementary lever to model scaling.

Harness design for long-running application development

Anthropic Engineering

Anthropic engineers detail a multi-agent harness design using generator and evaluator agents to improve Claude's ability to build complete, high-quality frontend applications autonomously over long durations.

@sethkarten: https://x.com/sethkarten/status/2072034978112889328

X AI KOLs Following

Continual Harness is a reset-free, self-improving agentic harness that achieves 20.54% on ARC-AGI-3 at a cost of $774 by storing memories, reusing skills, and refining its prompt, outperforming prior baselines like Hermes and OpenClaw with greater efficiency.