jaunt 0.4.1

Spec-driven code generation framework: write intent as decorated Python stubs, generate implementations and tests with LLMs.

Jaunt

_{William Blake, The Tyger, plate 42 from Songs of Experience (1794). The Metropolitan Museum of Art, Open Access.}

Tyger Tyger, burning bright, In the forests of the night; What immortal hand or eye, Could frame thy fearful symmetry?

-- William Blake, via Alfred Bester's The Stars My Destination

Jaunt is a small Python library + CLI for spec-driven code generation:

• Write implementation intent as normal Python stubs decorated with @jaunt.magic(...).
• Optionally write test intent as stubs decorated with @jaunt.test(...).
• Jaunt generates real modules under __generated__/ using an LLM backend (OpenAI, Anthropic, or Cerebras).
• Async support is available for both implementation and test specs through async def plus the build.async_runner setting.
• @magic works on individual class methods too — decorate instance methods, @classmethod, @staticmethod, or @abstractmethod stubs and Jaunt generates only those methods while preserving the rest of the class.
• Incremental freshness tracks both module digests and exported dependency APIs, so signature changes, whole-class member changes, and contract docstring edits can invalidate dependents.

Installation

pip install jaunt[openai]      # OpenAI + default Aider runtime
pip install jaunt[anthropic]   # Anthropic/Claude + default Aider runtime
pip install jaunt[cerebras]    # Cerebras + default Aider runtime
pip install jaunt[aider]       # runtime-only extra for custom install setups
pip install jaunt[all]         # all bundled backends/tools

Aider Runtime

Jaunt now defaults to agent.engine = "aider" for its internal build/test/skill workflows. Set agent.engine = "legacy" if you want the older direct-SDK runtime instead.

Practical limitation today: if you use Aider with a custom llm.api_key_env name that differs from the provider's canonical variable (OPENAI_API_KEY, ANTHROPIC_API_KEY, CEREBRAS_API_KEY), Jaunt currently remaps that key through os.environ under a global lock. That keeps auth stable, but it serializes concurrent Aider tasks for that config. For full parallelism today, prefer the canonical provider env var name.

Quickstart (This Repo)

Prereqs: uv installed.

uv sync
export OPENAI_API_KEY=...   # or ANTHROPIC_API_KEY for Claude
uv run jaunt --version

See docs-site/ for rendered docs, or DOCS.md for a plain-text walkthrough.

All examples live under examples/. See examples/README.md for the full list.

Hackathon Demo (JWT Auth)

Headline demo: JWT auth (the "wow gap" example: short spec, real generated glue + tests).

# Generate implementations for @jaunt.magic specs.
uv run jaunt build --root examples/jwt_auth

# Generate pytest tests for @jaunt.test specs and run them.
PYTHONPATH=examples/jwt_auth/src uv run jaunt test --root examples/jwt_auth

Freshness Model

• The full cleaned docstring is part of the spec contract, not just the first summary line.
• For whole-class @jaunt.magic specs, Jaunt treats the class signature plus declared members and method signatures as exported API.
• Jaunt's freshness model uses that dependency API too, so an upstream contract change can mark downstream modules stale even if their own source file did not change.

Eval Suite

Run the built-in eval suite against your configured backend:

uv run jaunt eval
uv run jaunt eval --model gpt-4o
uv run jaunt eval --provider anthropic --model claude-sonnet-4-5-20250929

Compare explicit provider/model targets:

uv run jaunt eval --compare openai:gpt-4o anthropic:claude-sonnet-4-5-20250929

Eval outputs are written under .jaunt/evals/<timestamp>/.

Eval Results (2026-02-15 UTC)

Run (UTC)	Mode	Target	Reasoning	Passed	Failed	Skipped	Total	Notes	Artifacts
2026-02-15T21-34-58Z	single	cerebras:gpt-oss-120b	none	0	10	0	10	Missing cerebras-cloud-sdk dependency	examples/expr_eval/.jaunt/evals/2026-02-15T21-34-58Z
2026-02-15T21-35-17Z	single	cerebras:gpt-oss-120b	none	0	10	0	10	Cerebras 402 payment_required quota/billing error	examples/expr_eval/.jaunt/evals/2026-02-15T21-35-17Z
2026-02-15T21-36-54Z	single	cerebras:gpt-oss-120b	none	10	0	0	10	All eval cases passed	examples/expr_eval/.jaunt/evals/2026-02-15T21-36-54Z
2026-02-15T22-01-24Z-custom-compare	compare	cerebras:gpt-oss-120b	low	10	0	0	10	All eval cases passed	examples/expr_eval/.jaunt/evals/2026-02-15T22-01-24Z-custom-compare
2026-02-15T22-01-24Z-custom-compare	compare	openai:gpt-5.2	none	10	0	0	10	All eval cases passed	examples/expr_eval/.jaunt/evals/2026-02-15T22-01-24Z-custom-compare
2026-02-15T22-01-24Z-custom-compare	compare	anthropic:opus-4.6	none	0	10	0	10	Anthropic 404 not_found_error for model name	examples/expr_eval/.jaunt/evals/2026-02-15T22-01-24Z-custom-compare
2026-02-15T22-04-19Z-custom-compare	compare	cerebras:gpt-oss-120b	low	10	0	0	10	All eval cases passed	examples/expr_eval/.jaunt/evals/2026-02-15T22-04-19Z-custom-compare
2026-02-15T22-04-19Z-custom-compare	compare	openai:gpt-5.2	none	10	0	0	10	All eval cases passed	examples/expr_eval/.jaunt/evals/2026-02-15T22-04-19Z-custom-compare
2026-02-15T22-04-19Z-custom-compare	compare	anthropic:claude-haiku-4-5	none	9	1	0	10	One assertion failure (example_slugify_smoke)	examples/expr_eval/.jaunt/evals/2026-02-15T22-04-19Z-custom-compare

Prompt snapshots:

uv run pytest tests/test_prompt_snapshots.py --snapshot-update

Auto-Generate PyPI Skills (Build)

jaunt build includes a best-effort pre-build step that auto-generates “skills” for external libraries your project imports and injects them into the build prompt.

What happens:

• Scan paths.source_roots for import ... / from ... import ... (ignores stdlib, internal modules, and relative imports).
• Resolve imports to installed PyPI distributions + versions from the current environment.
• Ensure a skill exists per distribution at:
  • <project_root>/.agents/skills/<dist-normalized>/SKILL.md
• If missing/outdated, fetch the exact PyPI README for <dist>==<version> and generate SKILL.md using the configured LLM provider.
• Inject the concatenated skills text into the build LLM prompt.

Overwrite rules:

• Jaunt only overwrites a skill if it was previously Jaunt-generated (it has a <!-- jaunt:skill=pypi ... --> header) and the installed version changed.
• If the header is missing, the file is treated as user-managed and will never be overwritten.

Failure mode: warnings to stderr, and the build continues without missing skills.

Docs Site (Fumadocs)

The repository includes a Fumadocs (Next.js) documentation site under docs-site/.

cd docs-site
npm run dev

Publish to PyPI

If you keep your token in .env as UV_PUBLISH_TOKEN=..., load it into your shell first:

set -a
source .env
set +a

Build and validate artifacts:

uv build
uvx twine check dist/*

Upload to PyPI:

uv publish --check-url https://pypi.org/simple/

Dev

uv run ruff check --fix .
uv run ruff format .
uv run ty check
uv run pytest

Final verification before pushing:

uv run ruff check .
uv run ruff format --check .
uv run ty check
uv run pytest

Why "Jaunt"?

Named after jaunting -- teleportation by thought alone -- from Alfred Bester's 1956 novel The Stars My Destination (originally published as Tiger! Tiger!). You think about where you want to be, and you're there.

Jaunt works the same way: describe your intent, and arrive at working code.

The forge-and-furnace imagery you'll find scattered through the codebase comes from William Blake's poem "The Tyger," which Bester used as the novel's epigraph and alternate title. The poem's vision of creation -- hammer, chain, furnace, anvil -- mirrors the act of forging code from pure specification.