Weasel Quickstart

chimera weasel Quickstart

chimera weasel is the fourth Chimera coding-agent CLI. Where chimera mink mirrors a TUI-first agent, chimera otter mirrors a server-first multi-client agent, and chimera ferret mirrors an IDE-first sandbox-first agent, weasel mirrors the minimal harness: powerful defaults plus four operating modes, adapt-to-your-workflow rather than ship-every-feature.

The headline trade is simplicity. Weasel ships no sub-agents, no plan mode, no built-in approval presets, no opinionated session chrome. What it ships is a clean four-mode entry surface, an auto-discovered .weasel/extensions/ directory, and an embeddable Agent class. If you want more, you build it (or install an extension); weasel will not get in the way.

This page walks the four entry points end-to-end. For deeper dives:

• modes.md — interactive / print / rpc / sdk in detail.
• extensions.md — .weasel/extensions/ layout.
• sdk.md — from chimera.weasel.sdk import Agent.
• providers.md — provider chain.
• parity-matrix.md — upstream surface mapping.
• security-and-trademarks.md — policy.

Prerequisites

• Python 3.11+
• uv
• One of: an Anthropic API key, an OpenAI API key, an OpenRouter API key, or a running Ollama daemon

uv --version                          # >= 0.4
uv sync --extra dev --extra anthropic # core + Anthropic SDK

The Anthropic extra is recommended for the default model. To drive weasel against OpenAI, OpenRouter, llama.cpp, or Ollama, swap the extra (--extra openai) or skip it; the OpenAI-compatible adapter is stdlib + httpx. The full matrix lives in providers.md.

Provider configuration

Weasel resolves the provider in this order (first match wins):

1. --model <id> on the CLI.
2. $WEASEL_MODEL environment variable.
3. $ANTHROPIC_API_KEY set → defaults to claude-sonnet-4-6.
4. $OPENAI_API_KEY set → defaults to gpt-4o.
5. $OPENROUTER_API_KEY set → defaults to anthropic/claude-sonnet-4.
6. Local Ollama daemon reachable on :11434 → first installed tag.
7. Friendly error pointing at the env vars above.

export ANTHROPIC_API_KEY=sk-ant-...
# OR
export OPENAI_API_KEY=sk-...
# OR
export OPENROUTER_API_KEY=sk-or-...

The four modes at a glance

Mode	Command	When to use
Interactive	chimera weasel	Day-to-day, conversational, mid-turn steering.
Print (one-shot)	chimera weasel -p "..."	Scripts, CI, xargs, shell pipelines.
RPC (stdio)	chimera weasel --mode rpc	Process integration: another tool drives weasel over JSON-RPC.
SDK (embedded)	from chimera.weasel.sdk import Agent	Drop into your Python app; no subprocess.

The modes share one loop, one tool registry, one extension surface, and one session store. Switching modes does not change semantics — only the I/O envelope.

Mode 1 — Interactive REPL

Run chimera weasel with no flags for a streaming REPL:

chimera weasel

weasel · claude-sonnet-4-6 · /Users/me/proj
> list the top-level files and read the README
I'll list the repo first, then read the README.

▶ list_files(path=".")
CHANGELOG.md  CLAUDE.md  README.md  chimera/  docs/  examples/  tests/

▶ Read(path="README.md")
# Chimera
A composable coding agent framework
...

The repo root has a README pitching Chimera as a composable coding agent framework.
> ▌

The REPL streams assistant text + tool calls inline, accepts mid-turn steering (just type while the agent is working), and supports Ctrl-C cancellation. Slash commands are intentionally sparse: /help, /exit, /model, /cost, /clear, /sessions, /extensions. Anything else you want, you add via an extension.

Every REPL session is event-sourced under ~/.chimera/eventlog/weasel-<utc>-<uuid>/. To resume:

chimera weasel sessions list
chimera weasel sessions show weasel-20260430T101455-1f3c2a8b

Mode 2 — Print (one-shot)

-p runs a single turn and exits. Plain text on stdout by default, JSON with --json:

chimera weasel -p "list the top-level files and read the README"
chimera weasel -p "summarize TODO comments in src/" --json
chimera weasel -p "ship it" --max-steps 5
chimera weasel --model gpt-4o -p "draft a release note"
chimera weasel -p "audit" --allowed-tools Read,Bash --no-save

Stdout is the agent’s final text answer (or one JSON blob with --json). Stderr carries the run-id banner so pipelines can keep stdout clean:

chimera weasel -p "what does this repo do" | tee summary.txt

Mode 3 — RPC (stdio JSON-RPC)

--mode rpc turns weasel into a JSON-RPC 2.0 server speaking on stdin/stdout — the integration point for outside tools that want to drive a weasel session as a subprocess.

chimera weasel --mode rpc < requests.jsonl

Request:

{"jsonrpc":"2.0","id":1,"method":"prompt","params":{"text":"list files"}}

Response stream:

{"jsonrpc":"2.0","method":"event","params":{"type":"text_delta","text":"I'll "}}
{"jsonrpc":"2.0","method":"event","params":{"type":"tool_call","name":"list_files"}}
{"jsonrpc":"2.0","id":1,"result":{"text":"...","cost":0.0042}}

Methods: prompt, steer, cancel, get_state, compact, list_sessions, resume. Full schema lives in modes.md.

Mode 4 — SDK (embedded)

For when you want weasel inside your Python process — no subprocess, no JSON-RPC, just a class. Both sync and async forms ship.

from chimera.weasel.sdk import Agent

agent = Agent(model="claude-sonnet-4-6")
result = agent.run("list the top-level files and read the README")
print(result.text)
print(f"cost: ${result.cost:.4f}")

Async form:

import asyncio
from chimera.weasel.sdk import Agent

async def main() -> None:
    agent = Agent(model="claude-sonnet-4-6")
    async for event in agent.stream("explain this repo"):
        if event.type == "text_delta":
            print(event.text, end="", flush=True)

asyncio.run(main())

Full recipe lives in sdk.md.

Extensions

Weasel auto-discovers .weasel/extensions/*.{py,js,ts} in the cwd and ~/.weasel/extensions/ globally. Extensions register tools, hooks, slash commands, and prompt templates. Layout:

.weasel/
  settings.json
  extensions/
    my-tool.py            # Python (importlib)
    my-script.ts          # TS/JS (subprocess via Node)
    feature/
      manifest.json
      index.py

A minimal Python extension:

from chimera.weasel.sdk import extension, tool

@extension(name="hello", version="0.1")
def register(api):
    @tool
    def hello(name: str) -> str:
        """Say hi."""
        return f"hello, {name}!"
    api.register_tool(hello)

Drop it under .weasel/extensions/hello.py and the next chimera weasel invocation will pick it up. Full schema in extensions.md.

Env vars at a glance

Variable	Default	Meaning
WEASEL_MODEL	(unset)	Default model id when --model is not passed.
ANTHROPIC_API_KEY	(unset)	Activates the Anthropic provider chain.
OPENAI_API_KEY	(unset)	Activates the OpenAI provider chain.
OPENROUTER_API_KEY	(unset)	Activates the OpenRouter (compatible) provider chain.
OLLAMA_HOST	http://localhost:11434	Ollama daemon root (set for remote daemons).
WEASEL_EXTENSIONS_DIR	.weasel/extensions/	Override the extensions search root.
NO_COLOR	(unset)	Force the plain output handler.

What gets written to disk

Path	What
~/.chimera/eventlog/weasel-<id>/	Per-run event stream + summary.
.weasel/settings.json	Project-local settings (model, extensions allowlist).
.weasel/extensions/	Project-local extensions.
~/.weasel/extensions/	User-global extensions.
~/.chimera/credentials.json	OAuth-issued tokens (mode 0o600).

Everything is local and plaintext. No remote telemetry. To purge old runs: rm -rf ~/.chimera/eventlog/weasel-*.

Where to go next

• Pick the right mode for your job: modes.md.
• Write an extension: extensions.md.
• Embed weasel in a Python app: sdk.md.
• Pick a provider: providers.md.
• Want the surface-by-surface parity status? parity-matrix.md.
• Trademark + security policy: security-and-trademarks.md.