> ## Documentation Index
> Fetch the complete documentation index at: https://docs.harborframework.com/llms.txt
> Use this file to discover all available pages before exploring further.

# News

<div id="news-page" />

<Update label="September 16, 2026">
  ## New Harbor docs

  Harbor docs now cover previously undocumented features such as simulated users,
  streaming, and regrading, and expose an MCP for coding agents.
  [Read more.](https://x.com/alexgshaw/status/2100296774955237438)
</Update>

<Update label="September 14, 2026">
  ## Watch agents as they run

  Harbor streams the agent’s trajectory and sandbox files while a trial is in progress.
  [Read more.](https://x.com/kobe0938/status/2099537332462596436)
</Update>

<Update label="August 29, 2026">
  ## Terminal-Bench 4.0

  Terminal-Bench 4.0 calibrates task resources, applies task fixes, and removes
  saturated tasks on Harbor Hub.
  [Read more.](https://x.com/ryan_marten/status/2093523335972036657)
</Update>

<Update label="August 27, 2026">
  ## Terminal-Bench-Science

  A Stanford-led Harbor benchmark for scientific research workflows: 70 tasks
  across life, physical, mathematical, earth, and engineering sciences, with
  Opus 5 at about 30%.
  [Read more.](https://x.com/StevenDillmann/status/2093041660615852448)
</Update>

<Update label="August 25, 2026">
  ## Simulated users

  A user agent can role-play a human and drive the target agent through a
  multi-turn conversation over ACP.
  [Read more.](https://x.com/kobe0938/status/2092298110391472430)
</Update>

<Update label="August 13, 2026">
  ## Regrade without rerunning the agent

  `harbor regrade` runs a new grader on recorded trials so you can iterate on
  verifiers without paying for another agent session.
  [Read more.](https://x.com/kobe0938/status/2087941751470170411)
</Update>

<Update label="August 11, 2026">
  ## Load a previous trajectory into a new eval

  `--load-trajectory` seeds a Harbor run with a prior Claude Code, Codex, or ATIF
  session so context moves into the next trial.
  [Read more.](https://x.com/kobe0938/status/2087273674818920768)
</Update>

<Update label="August 10, 2026">
  ## Interview an agent after the eval

  `harbor trial handoff` pulls a finished trial into local Claude Code so you can
  inspect the work and ask follow-up questions.
  [Read more.](https://x.com/kobe0938/status/2086905724295422008)
</Update>

<Update label="July 23, 2026">
  ## Terminal-Bench 3.0 (formerly Frontier-Bench)

  A living 74-task benchmark from the Harbor team where the best agents scored
  about 34%, runnable from the Harbor CLI and Hub.
  [Read more.](https://x.com/ryan_marten/status/2080322620248281252)

  ## Continuous Benchmarks

  Benchmarks are not static artifacts. They are software and we should maintain
  them like software.
  [Read more.](https://x.com/ryan_marten/status/2080321791361527843)
</Update>

<Update label="July 9, 2026">
  ## Hosted leaderboards on Harbor Hub

  Publish dataset leaderboards on Harbor Hub, including reward-hacking detection
  on Terminal-Bench 2.1.
  [Read more.](https://x.com/alexgshaw/status/2075273684424790516)
</Update>

<Update label="July 8, 2026">
  ## Harbor-Index

  Compose the hardest Harbor tasks into an 82-task meta-benchmark that no agent
  cleared above 30% at launch.
  [Read more.](https://x.com/LinShi592021/status/2074722423568736717)
</Update>

<Update label="July 3, 2026">
  ## Introducing harbor exec

  Run an agentic map-reduce over sandboxes to analyze traces, mine sessions, and
  search across completed jobs.
  [Read more.](https://x.com/alexgshaw/status/2073086868565196946)
</Update>

<Update label="June 18, 2026">
  ## Sidecar environments

  Collect artifacts from sidecars, run verification in a sidecar, and execute
  pre-verification commands in any service.
  [Read more.](https://x.com/harborframework/status/2067656083825373652)
</Update>

<Update label="June 17, 2026">
  ## Evaluate ACP agents natively

  Harbor can run ACP registry agents with `harbor run --agent acp:junie`, adding
  36 agents now and every future ACP agent when it lands in the registry.
  [Read more.](https://x.com/harborframework/status/2067314067803701255)
</Update>

<Update label="May 27, 2026">
  ## Stop zipping your job results

  Upload and share Harbor job results on Harbor Hub instead of zipping and
  sending them manually.
  [Read more.](/news/job-result-sharing)
</Update>

<Update label="May 15, 2026">
  ## Verify Harbor tasks in a separate sandbox

  Run task verification in a sandbox separate from the agent, with explicit
  artifact handoff between the two environments.
  [Read more.](/news/separate-verifier-sandboxes)
</Update>

<Update label="April 23, 2026">
  ## Multi-step tasks

  The first major addition to the Harbor task format: tasks split into sequential
  steps, each with its own instruction and verifier.
  [Read more.](/news/multi-step-tasks)
</Update>

<Update label="March 27, 2026">
  ## Explore the Harbor cookbook

  Recipes for building Harbor tasks and optimization loops.
  [Read more.](/news/harbor-cookbook)

  ## The Harbor registry is getting an upgrade

  A new way to package and distribute Harbor tasks and datasets.
  [Read more.](/news/harbor-registry)
</Update>