Skip to main content

AI coding orchestrator with self-improving specialist agents

Project description

Wagner CLI

AI coding orchestrator with self-improving specialist agents.

Wagner decomposes complex coding tasks across specialist agents, each trained on domain-specific patterns. The more you use it, the smarter it gets.

Install

pip install wagner-cli

Quick Start

Wagner runs on a bring-your-own-key model: orchestration (decomposition, routing, specialist faces) runs on the Wagner server, while every model call runs on your Anthropic key. Your key is sent per request and is never stored or logged server-side.

# 1. Point Wagner at your own Anthropic key (stored in ~/.wagner/config.json)
wagner config --anthropic-key sk-ant-...

# 2. Orchestrate a coding task
wagner run "build a REST API with JWT auth and PostgreSQL"

# 3. Write the generated files to disk
wagner run "build a CLI tool that watches files" --write

More examples:

# Dry run — see the decomposition without synthesis (no synthesis cost)
wagner run "build a chat app" --dry

# Skip clarification / critic passes
wagner run "build a FastAPI CRUD app" --no-clarify --no-critic

# Edit an existing codebase — pass relevant files as context
wagner run "add pagination to the list endpoint" \
  --context app/routes.py app/models.py --write

Commands

wagner run "prompt"      Orchestrate a coding task
wagner config            Configure your Anthropic key / server
wagner status            Show server status and face count
wagner faces             List all specialist faces
wagner history           Show recent orchestrations
wagner gaps              Show missing specialist domains

wagner run options

--write, -w        Write generated files to the output directory
--output, -o DIR   Output directory (default: current directory)
--overwrite        Overwrite existing files
--dry              Decompose only, no synthesis
--no-clarify       Skip clarification questions
--no-critic        Skip the critic/revision pass
--context, -c PATH Source file to include as context (repeatable)
--server, -s URL   Override the server for this command

Configuration

# Set your Anthropic key (BYOK)
wagner config --anthropic-key sk-ant-...

# Use a custom server
wagner config --anthropic-key sk-ant-... --server https://wagner.razzyshmazzy.com

Any flag you omit keeps its current value. Config lives in ~/.wagner/config.json.

How It Works

  1. You type a prompt
  2. Wagner's Clarifier checks if the prompt is ambiguous
  3. The Decomposer breaks it into specialist subtasks
  4. Each subtask is dispatched to the best specialist Face
  5. The Synthesizer combines outputs into structured files
  6. The Critic reviews and triggers revision if needed
  7. Files are written to disk with --write

Every prompt improves the system. Faces accumulate examples and get smarter over time.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wagner_cli-0.1.0.tar.gz (10.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

wagner_cli-0.1.0-py3-none-any.whl (10.7 kB view details)

Uploaded Python 3

File details

Details for the file wagner_cli-0.1.0.tar.gz.

File metadata

  • Download URL: wagner_cli-0.1.0.tar.gz
  • Upload date:
  • Size: 10.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for wagner_cli-0.1.0.tar.gz
Algorithm Hash digest
SHA256 490ecca42afdce71ce87930c8e596ae8963a02ef007902b0bc7eab55433f863b
MD5 b07367dcafa12252e6996b03d36dfcee
BLAKE2b-256 5ff42d1ca50c9f26499d735aa231934ba5d5d484306c60f6152eabc53664e49b

See more details on using hashes here.

File details

Details for the file wagner_cli-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: wagner_cli-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 10.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for wagner_cli-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 88b1386c01fb797df339d7d001b610edfb79569f9f44563c352847e3331b3fb4
MD5 ad523f60e6805928716a599599e7525e
BLAKE2b-256 365a18c9d06067295a1373e3f0bf8c0c695b7cc1ee2f5e3250b1ba9d305df10c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page