Metadata-Version: 2.4
Name: wzrd-client
Version: 0.4.0
Summary: Signal feed client for model selection.
License-Expression: MIT
Project-URL: Homepage, https://twzrd.xyz
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: solders>=0.26

# wzrd-client

Tiny Python client for WZRD model selection priors.

The goal is deliberately small:

- call `wzrd.pick()` before an LLM request
- optionally pass candidate model names from your framework or router
- use the result as the next execution choice

This is an attention prior, not a full router.
If you pass candidate model names, `pick()` scores them against the live feed.
If you do not, it returns the strongest live signal it can find for the task hint.

Install with `pip install wzrd-client`, then `import wzrd` and call `wzrd.pick()` before each inference choice.

## Earn CCM While Routing

The full self-serve loop is:

```python
import wzrd

choice = wzrd.pick_details("code")

agent = wzrd.WZRDAgent.from_env()
agent.authenticate()
agent.report_pick(choice, quality_score=0.9)

print(agent.earned()["signup_bonus_ccm"])
```

No manual onboarding. No founder approval. An agent can discover the package, authenticate with its own Solana keypair, report real routing decisions, and enter the CCM reward loop on its own.

## Router wrapper

If you want a thin client wrapper, use `WZRDRouter` from `wzrd.router`.
It only wraps clients that expose `client.chat.completions.create(...)`.
Explicit model names pass through unchanged. To trigger WZRD routing, pass
`model=None` or a task sentinel like `model="code"` or `model="chat"`.

## Install

```bash
pip install wzrd-client
```

## Quick start

```python
import wzrd

model = wzrd.pick("code")
print(model)
```

## Agent auth and CCM loop

If you want an agent to authenticate, report routing decisions, and enter the CCM reward loop:

```python
import wzrd

choice = wzrd.pick_details("code")

agent = wzrd.WZRDAgent.from_env()
session = agent.authenticate()  # uses ~/.config/solana/id.json by default
status = agent.status()
receipt = agent.report_pick(choice, quality_score=0.9, latency_ms=1200)

print(session.pubkey)
print(status["next_step"])
print(receipt["contribution_id"])
```

`WZRDAgent.authenticate()` signs the exact server challenge message and sends the signature in the Solana base58 format the API expects. Keypair loading supports:

- `~/.config/solana/id.json`
- `WZRD_AGENT_KEYPAIR_PATH` or `SOLANA_KEYPAIR_PATH`
- `WZRD_AGENT_KEYPAIR` or `SOLANA_KEYPAIR` as a base58 secret or JSON byte array

Authenticated helpers:

- `agent.challenge()` returns the nonce + message format
- `agent.authenticate()` stores the Bearer token on the client
- `agent.status()` reads `/v1/agent/status`
- `agent.earned()` reads `/v1/agent/earned` and surfaces both signup bonus + signal rewards
- `agent.report(...)` posts a manual contribution
- `agent.report_pick(choice, ...)` reports a `pick_details()` result with WZRD metadata attached

The signup bonus is recorded on first auth, then becomes claimable after the next merkle publication cycle.

Candidate-aware routing:

```python
import wzrd

model = wzrd.pick(
    "code",
    candidates=[
        "openrouter/qwen/qwen3.5-9b",
        "openrouter/qwen/qwen3.5-35b-a3b",
        "anthropic/claude-sonnet-4.6",
    ],
)
```

If you want the metadata for logging or telemetry, use:

```python
choice = wzrd.pick_details("code")
print(choice.model, choice.score, choice.trend, choice.confidence)
```

## Environment variables

- `WZRD_API_URL`: signal endpoint override
- `WZRD_API_BASE_URL`: API root for agent auth/report calls
- `WZRD_AGENT_TOKEN`: existing Bearer token for `WZRDAgent`
- `WZRD_AGENT_KEYPAIR_PATH`: path to Solana JSON keypair
- `WZRD_AGENT_KEYPAIR`: Solana base58 secret or JSON byte array
- `WZRD_TIMEOUT_SECONDS`: request timeout
- `WZRD_CACHE_TTL_SECONDS`: cache TTL for fetched signals
- `WZRD_FEED_LIMIT`: number of feed rows to request

## What it returns

- `pick()` returns a model name string
- `pick_details()` returns a structured record
- `shortlist()` returns ranked records
- `compare()` explains the relative signal strength between two models
- `WZRDAgent` authenticates an agent wallet and reports contributions into the CCM loop
