Open Source Radar
taxideftis/

bathos

GitHub

Bathos is a Rust engine-based AI workflow agent with a 7-wave pipeline, 17 roles, and scale-adaptive routing. It includes CLI references, installation/build steps, and platform-specific setup instructions in the README.

16stars
8forks
0issues
2026since
Star historydaily snapshots by VibeCrowd

Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (2 so far).

Reviewgenerated from repository data · Jul 18, 2026

What it is

BATHOS is an AI workflow agent consisting of a small Rust engine that enforces invariants and gates within a structured delivery pipeline. It presents a 17 specialist roles framework, a 7-wave pipeline, scale-adaptive routing, and hard quality gates, backed by a Rust executable named bathos.

How it works

BATHOS operates on two planes: orchestration (Markdown slash commands, roles, hooks under .claude/) and a Rust engine (a single static binary, bathos) that enforces state, gates, wave transitions, routing, and story freshness. The engine is invoked by hooks and commands such as bathos <subcommand>. The README includes examples of using bathos with commands like /wave1-discovery and with a JSON state input to route decisions.

Getting started

Getting started involves cloning the repository, building the engine, and wiring the binary for hook discovery:

  • macOS/Linux build path shows building the core engine and setting BATHOS_BIN, with commands like:
git clone <your-fork-url> bathos && cd bathos

# Build the single static engine binary (~5.9 MB)
cdd core
cargo build --release
cd ..

# Make the engine discoverable by hooks/commands:
export BATHOS_BIN="$(pwd)/core/target/release/bathos"
  • Windows (PowerShell) path shows running install.ps1 and setting BATHOS_BIN to bathos.exe, with commands:
.\install.ps1
  • WSL path indicates using the Linux path and building the Linux binary, with environment setup similar to macOS/Linux.

Recent releases

RELEASES (latest 0): - none

Traction

Stars: 16

Behind the repo

No startup/company details provided in the README beyond project scope and tooling description.

Caveats

License: MIT badge shown in README, but the repo metadata states license: none listed. Prerequisites include Claude Code v2.1.32+ with Agent Teams, Rust toolchain (cargo 1.92+), jq (macOS/Linux), and appropriate shell or PowerShell setup depending on platform. Windows wiring changes involve copying settings.json variants and hook scripts.

SharePost on XLinkedIn
All trending reposRevenue-verified startups →