Topic hub · 13 pieces

Agent harnesses and runtimes

Mitchell Hashimoto used the term harness engineering on February 5, 2026, and Viv Trivedy gave it a formal taxonomy twelve days later. By then twelve coding agents, from Claude Code to OpenCode, had built the same harness primitives independently, hooks among them.

The value of that layer now has a number. A controlled study found harness-induced variance 7.8 times larger than model-induced variance on a SWE-bench subset, and one model moved from 19.1% to 73.4% on Claw-SWE-Bench with only the adapter changed. Underneath, 23 companies sell agents a sandboxed computer, and Modal raised $355 million at $4.65 billion on May 21.

Timeline

  1. · Agents

    Inside the personal AI runtime wars: 487K stars, 57K issues, 25 SaaS businesses

    OpenClaw has 138 CVEs, a 1GB/min memory leak and a supply-chain attack where 80% of audited skills were malicious; Hermes's flagship feature silently deleted user data on launch day; and 25+ SaaS providers now live off that pain.

  2. · Tools

    Flue, Fred Schott, and the agent-harness moment in TypeScript

    Fred Schott's Flue treats the agent harness as a first-class TypeScript build target; this update covers the 1.0 Beta rewrite through beta.4, the Actions primitive, @flue/react, the Cloudflare runtime deal and the Vercel-backed eve rival.

  3. · Tools

    Agent harness engineering: the discipline nobody named for two years

    Two near-simultaneous coiners (Mitchell Hashimoto and Viv Trivedy), seven convergent voices and twelve coding agents that built the same primitives independently mark a discipline that crystallized in nine months.

  4. · Tools

    The agent-sandbox wars: 23 companies and a contested substrate

    Modal raised $355M at $4.65B selling AI agents a computer, as do 22 other companies; this map of the sandbox race shows who runs Firecracker, who doesn't, and why the substrate war is nearly settled while everything above it is wide open.

  5. · Agents

    The agent-framework field reinvents the same five wheels

    Eleven OSS agent runtimes share one recurring pattern of memory, sandboxing, skills-vs-MCP, tool governance, and transport: only Letta solved memory, only smolagents solved sandboxing, and the 2026 vendor SDKs are quietly absorbing the rest into hosting platforms.

  6. · Tools

    Loop engineering and the terminology treadmill: what 47 Claude Code runs actually show

    On June 7, 2026, X renamed the agent loop 'loop engineering' and called it the successor to harness engineering; The Deep Feed ran 47 Claude Code loops in a sandbox and found the naming churns while three of four technical claims hold, one to the cent.

  7. · Agents

    Who owns the coding-agent runtime?

    In 48 hours the question of where long-running coding agents actually run got both answers at once: an indie $7M bet against lock-in, and a frontier-lab acquisition that pulls the runtime layer inside Codex.

  8. · Tools

    SuperServe, field-tested: what 356 live assertions reveal about an agent runtime

    SuperServe's landing page makes four promises; The Deep Feed drove the live API 356 ways across 24 test phases — booting every template, running nine agent runtimes, executing a multi-agent swarm — to map the gap between the spec sheet and the substrate.

  9. · Tools

    Vercel's eve and the agent-as-a-directory bet

    Vercel open-sourced eve at Ship London, a framework where an agent is a folder of files; The Deep Feed built 75 agents on it and ran an adversarial review to find what holds up, what breaks, and the stateful-on-stateless tension the launch posts skip.

  10. · Agents

    How much is the harness worth? The year the number got measured

    Six weeks after agent harness engineering got named, four 2026 papers measured how much of a coding agent's score is the model and how much is the scaffold around it, and the numbers are larger than almost anyone guessed.

  11. · Tools

    Mastra in production: the advanced patterns worth shipping

    An operator playbook for Mastra across six TypeScript codebases and the full 1.x line, covering memory's four layers, supervisors, MCP on both ends, RAG and evals, auth, observability, the 1.x migrations and the production bugs that taught the most.

  12. · Agents

    The Cursor SDK and the year the coding agent stopped being an editor

    Notion embedded coding agents into its workspace in a few weeks with the Cursor SDK, which turns the harness behind Cursor into a callable runtime so an @mention can plan, build, test and open a PR, a bigger shift than any model release.

  13. · Agents

    Always-on agents hand approvals to a second model and liability to the user

    Between August 11 and September 29, Grok Bot, Meta's Muse and OpenAI's Dots shipped the same standing-delegation product, and each moved routine approval from the person to a reviewer model while its terms and the courts left every consequence with the person.

In the data

Rows from the CC BY 4.0 datasets that these pieces report. Each links to its row, primary source and method.

Model releases · 1

Funding and acquisitions · 5

Key primary sources

More on this subject

Tag: harness Tag: agent-runtime Tag: sandbox Tag: agent-infrastructure Department: Agents Department: Tools All topic hubs