SuperQodeSUPERQODE
Menu
← Back to the Hub

DeepAgents

deepagents

Optional

LangChain DeepAgents coding harness over the DeepAgents SDK; DeepAgents owns the graph, filesystem backend, and subagents.

PythondeepagentsOpen sourceFresh session sessions
Repository langchain-ai/deepagents ↗Documentation docs.superqode.dev ↗Harness Hub guide

DeepAgents runs its own filesystem and execute tools, so SuperQode approvals do not gate them. Model routes resolve through LangChain's provider strings.

Specification

CategoryOptional integrations
LanguagePython
How we knowlangchain-ai/deepagents carries pyproject.toml (Python 98%).
Runtimedeepagents
InterfaceTUI and CLI
Sourceoptional:deepagents
LicenceMIT
ContinuityFresh session
Based ondeepagents

Runtime behaviour

Capabilities

single workflowStructured run events

Tools

readwriteeditbash

Policies and ownership

  • +Sandbox: local
  • +Approvals: none
  • +Shell: allowed
  • +Network: allowed
  • +Checks: not required

Get started

Inside SuperQode
$ :harness switch deepagents$ :harness show deepagents$ :harness status
From your shell
$ superqode harness run deepagents "Describe your task"$ superqode harness show deepagents

Install and authenticate

  1. 1

    Install the DeepAgents integration

    In the TUI, select Set up to install, verify, and continue without restarting.

    uv tool install "superqode[deepagents]"

Evaluate and optimize

Measure
$ superqode harness test --spec harness.yaml$ superqode harness eval --spec harness.yaml --tasks eval-tasks.yaml$ superqode harness bench --manifest harnessbench.yaml
Improve
$ superqode harness optimize-omni --spec harness.yaml --tasks eval-tasks.yaml$ superqode harness optimize --spec harness.yaml --tasks eval-tasks.yaml$ superqode harness promote stage

ALSO IN OPTIONAL INTEGRATIONS

RLM Coderlm-codeTau (Hugging Face)tauUHP harnessuhpDeepSeek Harnessdeepseek-harness
Browse all 83 harnesses