BenchRank
#21 in Code AssistantsUpdated 2026-08

Omnigent

by Omnigent · Open-source meta-harness for composing, governing and sharing AI agent sessions

42.8 — BenchRank score out of 100

Screenshots of Omnigent

  • Homepage

Homepage · Omnigent

Homepage of Omnigent

Overview

Omnigent is an open-source layer over AI agent harnesses — Claude Code, Codex, Pi or ones you write yourself. A runner wraps each agent in a sandboxed, uniform session that restricts filesystem and network access and hides credentials; a server adds policies such as spend caps and model routing, plus shared history. Sessions are reachable from the terminal, web, apps and a REST API.

Best for
Teams running several AI coding harnesses who want one layer for policies, sandboxing and shared sessions.
Pricing
No prices are shown on the captured page; the project is released under Apache 2.0 and installed by the user.
Runs on
macOSSelf-hostedCLI

Strengths and trade-offs

Strengths

  • One layer over Claude Code, Codex, Pi and your own agents
  • OS sandbox limits filesystem and network access, hides credentials
  • Sessions available from terminal, web, mobile and a REST API
  • Apache 2.0; installs via curl, uv, pip or Homebrew

Trade-offs

  • Described as alpha software on its own site
  • Custom agents and harnesses must be defined in YAML
  • No pricing, hosted option or support terms shown in the captured text

How this score is made up

Each dimension is scored out of 100 and combined into the headline score using fixed weights.

MCP support
0 out of 100
API quality
15 out of 100
Documentation
80 out of 100
Agent friendliness
60 out of 100
Changelog
100 out of 100
Marketing site structure
70 out of 100
Operational trust
0 out of 100

Measured, but not part of the score

Useful to know, but not a mark for or against the product — so these do not affect the ranking.

Openness
85 out of 100
Maintenance
100 out of 100

This doesn’t look right — report a problem with Omnigent’s score

Alternatives in Code Assistants

  • Ranked 1

    76.8 — BenchRank score out of 100

    Warp

    Warp · Terminal for running and orchestrating coding agents

    Best for: Developers running coding agents like Claude Code or Codex who want them managed in one terminal.

  • Ranked 2

    76 — BenchRank score out of 100

    Superset

    Superset · Desktop app for running coding agents in parallel Git worktrees

    Best for: Developers on macOS running several CLI coding agents in parallel across isolated Git worktrees

  • Ranked 3

    71.9 — BenchRank score out of 100

    Frontman

    Frontman · AI website editor for existing WordPress, Next.js, Astro and Vite sites

    Best for: Designers, PMs and marketing teams editing existing sites without waiting on developer tickets.

See all 42 alternatives to Omnigent

Report a problem with this page

Report an issue with Omnigent