Metadata-Version: 2.4
Name: avalanche-ai
Version: 0.1.0rc2
Summary: Python toolkit for local data-flow experiments with Iceberg and Lance
Author-email: Gabriel Lesperance <611342+glesperance@users.noreply.github.com>
License:                                  Apache License
                                   Version 2.0, January 2004
                                https://www.apache.org/licenses/
        
           TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
        
           1. Definitions.
        
              "License" shall mean the terms and conditions for use, reproduction,
              and distribution as defined by Sections 1 through 9 of this document.
        
              "Licensor" shall mean the copyright owner or entity authorized by
              the copyright owner that is granting the License.
        
              "Legal Entity" shall mean the union of the acting entity and all
              other entities that control, are controlled by, or are under common
              control with that entity. For the purposes of this definition,
              "control" means (i) the power, direct or indirect, to cause the
              direction or management of such entity, whether by contract or
              otherwise, or (ii) ownership of fifty percent (50%) or more of the
              outstanding shares, or (iii) beneficial ownership of such entity.
        
              "You" (or "Your") shall mean an individual or Legal Entity
              exercising permissions granted by this License.
        
              "Source" form shall mean the preferred form for making modifications,
              including but not limited to software source code, documentation
              source, and configuration files.
        
              "Object" form shall mean any form resulting from mechanical
              transformation or translation of a Source form, including but
              not limited to compiled object code, generated documentation,
              and conversions to other media types.
        
              "Work" shall mean the work of authorship, whether in Source or
              Object form, made available under the License, as indicated by a
              copyright notice that is included in or attached to the work
              (an example is provided in the Appendix below).
        
              "Derivative Works" shall mean any work, whether in Source or Object
              form, that is based on (or derived from) the Work and for which the
              editorial revisions, annotations, elaborations, or other modifications
              represent, as a whole, an original work of authorship. For the purposes
              of this License, Derivative Works shall not include works that remain
              separable from, or merely link (or bind by name) to the interfaces of,
              the Work and Derivative Works thereof.
        
              "Contribution" shall mean any work of authorship, including
              the original version of the Work and any modifications or additions
              to that Work or Derivative Works thereof, that is intentionally
              submitted to Licensor for inclusion in the Work by the copyright owner
              or by an individual or Legal Entity authorized to submit on behalf of
              the copyright owner. For the purposes of this definition, "submitted"
              means any form of electronic, verbal, or written communication sent
              to the Licensor or its representatives, including but not limited to
              communication on electronic mailing lists, source code control systems,
              and issue tracking systems that are managed by, or on behalf of, the
              Licensor for the purpose of discussing and improving the Work, but
              excluding communication that is conspicuously marked or otherwise
              designated in writing by the copyright owner as "Not a Contribution."
        
              "Contributor" shall mean Licensor and any individual or Legal Entity
              on behalf of whom a Contribution has been received by Licensor and
              subsequently incorporated within the Work.
        
           2. Grant of Copyright License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              copyright license to reproduce, prepare Derivative Works of,
              publicly display, publicly perform, sublicense, and distribute the
              Work and such Derivative Works in Source or Object form.
        
           3. Grant of Patent License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              (except as stated in this section) patent license to make, have made,
              use, offer to sell, sell, import, and otherwise transfer the Work,
              where such license applies only to those patent claims licensable
              by such Contributor that are necessarily infringed by their
              Contribution(s) alone or by combination of their Contribution(s)
              with the Work to which such Contribution(s) was submitted. If You
              institute patent litigation against any entity (including a
              cross-claim or counterclaim in a lawsuit) alleging that the Work
              or a Contribution incorporated within the Work constitutes direct
              or contributory patent infringement, then any patent licenses
              granted to You under this License for that Work shall terminate
              as of the date such litigation is filed.
        
           4. Redistribution. You may reproduce and distribute copies of the
              Work or Derivative Works thereof in any medium, with or without
              modifications, and in Source or Object form, provided that You
              meet the following conditions:
        
              (a) You must give any other recipients of the Work or
                  Derivative Works a copy of this License; and
        
              (b) You must cause any modified files to carry prominent notices
                  stating that You changed the files; and
        
              (c) You must retain, in the Source form of any Derivative Works
                  that You distribute, all copyright, patent, trademark, and
                  attribution notices from the Source form of the Work,
                  excluding those notices that do not pertain to any part of
                  the Derivative Works; and
        
              (d) If the Work includes a "NOTICE" text file as part of its
                  distribution, then any Derivative Works that You distribute must
                  include a readable copy of the attribution notices contained
                  within such NOTICE file, excluding those notices that do not
                  pertain to any part of the Derivative Works, in at least one
                  of the following places: within a NOTICE text file distributed
                  as part of the Derivative Works; within the Source form or
                  documentation, if provided along with the Derivative Works; or,
                  within a display generated by the Derivative Works, if and
                  wherever such third-party notices normally appear. The contents
                  of the NOTICE file are for informational purposes only and
                  do not modify the License. You may add Your own attribution
                  notices within Derivative Works that You distribute, alongside
                  or as an addendum to the NOTICE text from the Work, provided
                  that such additional attribution notices cannot be construed
                  as modifying the License.
        
              You may add Your own copyright statement to Your modifications and
              may provide additional or different license terms and conditions
              for use, reproduction, or distribution of Your modifications, or
              for any such Derivative Works as a whole, provided Your use,
              reproduction, and distribution of the Work otherwise complies with
              the conditions stated in this License.
        
           5. Submission of Contributions. Unless You explicitly state otherwise,
              any Contribution intentionally submitted for inclusion in the Work
              by You to the Licensor shall be under the terms and conditions of
              this License, without any additional terms or conditions.
              Notwithstanding the above, nothing herein shall supersede or modify
              the terms of any separate license agreement you may have executed
              with Licensor regarding such Contributions.
        
           6. Trademarks. This License does not grant permission to use the trade
              names, trademarks, service marks, or product names of the Licensor,
              except as required for reasonable and customary use in describing the
              origin of the Work and reproducing the content of the NOTICE file.
        
           7. Disclaimer of Warranty. Unless required by applicable law or
              agreed to in writing, Licensor provides the Work (and each
              Contributor provides its Contributions) on an "AS IS" BASIS,
              WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
              implied, including, without limitation, any warranties or conditions
              of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
              PARTICULAR PURPOSE. You are solely responsible for determining the
              appropriateness of using or redistributing the Work and assume any
              risks associated with Your exercise of permissions under this License.
        
           8. Limitation of Liability. In no event and under no legal theory,
              whether in tort (including negligence), contract, or otherwise,
              unless required by applicable law (such as deliberate and grossly
              negligent acts) or agreed to in writing, shall any Contributor be
              liable to You for damages, including any direct, indirect, special,
              incidental, or consequential damages of any character arising as a
              result of this License or out of the use or inability to use the
              Work (including but not limited to damages for loss of goodwill,
              work stoppage, computer failure or malfunction, or any and all
              other commercial damages or losses), even if such Contributor
              has been advised of the possibility of such damages.
        
           9. Accepting Warranty or Additional Liability. While redistributing
              the Work or Derivative Works thereof, You may choose to offer,
              and charge a fee for, acceptance of support, warranty, indemnity,
              or other liability obligations and/or rights consistent with this
              License. However, in accepting such obligations, You may act only
              on Your own behalf and on Your sole responsibility, not on behalf
              of any other Contributor, and only if You agree to indemnify,
              defend, and hold each Contributor harmless for any liability
              incurred by, or claims asserted against, such Contributor by reason
              of your accepting any such warranty or additional liability.
        
           END OF TERMS AND CONDITIONS
        
           APPENDIX: How to apply the Apache License to your work.
        
              To apply the Apache License to your work, attach the following
              boilerplate notice, with the fields enclosed by brackets "[]"
              replaced with your own identifying information. (Don't include
              the brackets!)  The text should be enclosed in the appropriate
              comment syntax for the file format. We also recommend that a
              file or class name and description of purpose be included on the
              same "printed page" as the copyright notice for easier
              identification within third-party archives.
        
           Copyright 2026 Trampoline AI Inc.
        
           Licensed under the Apache License, Version 2.0 (the "License");
           you may not use this file except in compliance with the License.
           You may obtain a copy of the License at
        
               https://www.apache.org/licenses/LICENSE-2.0
        
           Unless required by applicable law or agreed to in writing, software
           distributed under the License is distributed on an "AS IS" BASIS,
           WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
           See the License for the specific language governing permissions and
           limitations under the License.
License-File: LICENSE
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Requires-Python: <3.14,>=3.11
Requires-Dist: dataframely>=1.13.0
Requires-Dist: pydantic<3,>=2
Requires-Dist: pyiceberg[glue,pandas,pyarrow,s3fs,sql-sqlite]<0.11,>=0.9
Requires-Dist: python-ulid>=3.0.0
Provides-Extra: agent
Requires-Dist: predict-rlm>=0.8.0a; extra == 'agent'
Provides-Extra: all
Requires-Dist: croniter>=6.2.2; extra == 'all'
Requires-Dist: grpcio-tools>=1.75.1; extra == 'all'
Requires-Dist: grpcio>=1.75.1; extra == 'all'
Requires-Dist: predict-rlm>=0.8.0a; extra == 'all'
Requires-Dist: pylance>=0.24.0; extra == 'all'
Requires-Dist: ray[default]>=2.50.0; extra == 'all'
Requires-Dist: textual>=6.2.1; extra == 'all'
Requires-Dist: watchfiles>=1.1.1; extra == 'all'
Provides-Extra: lance
Requires-Dist: pylance>=0.24.0; extra == 'lance'
Provides-Extra: ray
Requires-Dist: ray[default]>=2.50.0; extra == 'ray'
Provides-Extra: runtime
Requires-Dist: croniter>=6.2.2; extra == 'runtime'
Requires-Dist: grpcio-tools>=1.75.1; extra == 'runtime'
Requires-Dist: grpcio>=1.75.1; extra == 'runtime'
Requires-Dist: watchfiles>=1.1.1; extra == 'runtime'
Provides-Extra: tui
Requires-Dist: grpcio>=1.75.1; extra == 'tui'
Requires-Dist: textual>=6.2.1; extra == 'tui'
Description-Content-Type: text/markdown

<p align="center">
  <img align="center" src="docs/assets/brand/avalanche-logo-3d.png" width="560px" alt="Avalanche logo" />
</p>


# Avalanche

Avalanche makes agents first-class steps in typed data pipelines. Compose  
adaptive agent work with deterministic Python transformations in one DAG, run  
it through the Avalanche operator, and inspect every run from the terminal UI.

> [!NOTE]  
> Avalanche is an early release candidate intended for local development and  
> experimentation. APIs and operational behavior may change before a stable  
> release.

## Installation

Add avalanche to your project:

```bash
uv add avalanche --all-extras
```

## Requirements

- Python 3.11, 3.12, or 3.13.
- Agent steps require the `agent` extra and credentials for the model provider
configured through PredictRLM/DSPy.
- The operator and connected TUI require the `runtime` and `tui` extras.
- Ray and Lance support are optional and require their corresponding extras.



## Quickstart



### With your coding agent

Give your coding agent Avalanche's workflow-authoring skill:

```bash
npx skills add Trampoline-AI/avalanche
```

### Workshop project setup

Create a project with editable local checkouts of Avalanche and PredictRLM:

```bash
npx skills add https://github.com/Trampoline-AI/avalanche/tree/main/.agents/skills/avalanche-demo-init
```



### Quick Example

```python
import random
import avalanche as ava

@ava.source
def generate_binary() -> str:
    length = random.randint(128, 256)
    return "1" + "".join(random.choice("01") for _ in range(length - 1))

@ava.agent_step(
    ava.Signature(
        "binary: str -> decimal: str",
    ),
    lm="openai/gpt-5.6-terra",
)
async def convert_binary(binary: str, *, agent: ava.Agent) -> str:
    return (await agent(binary=binary)).decimal

@ava.dest
def print_result(result: str) -> str:
    print(result)
    return result


@ava.workflow
def binary_converter():
    return generate_binary() >> convert_binary() >> print_result()
```

The handle is process-local and its non-daemon driver thread keeps the embedded
process alive until the run finishes. `run.cancel()` requests cooperative
cancellation between node submissions; it does not forcibly stop an active
thread or Ray task. Durable run state remains an operator responsibility.
Terminal values may contain `ava.File` or `ava.Workspace` directly or nested in
supported Pydantic/list/tuple/dict results. Embedded `.result()` returns the
original Python shape and portable file/workspace objects. A terminal
`Workspace` carries its serializable manifest; its local `.path` exists only
while Avalanche is executing user node code.

Start with the simplest smoke-tested example:
Run it through the local operator:

```bash
uv run ava dev --flows path/to/flow
```


## Real example agentic data transformation workflow

This example is closer to what a real production avalanche workflow can look like. This workflow turns product feedback into prioritized work in a CRM. It loads a typed corpus, runs theme and risk analysis in parallel, synthesizes both reports, and pushes the resulting plan to an external system.

Set the credential required by your model provider. For the model in this
example:

```bash
OPENAI_API_KEY="..."
ANTHROPIC_API_KEY="..."
GEMINI_API_KEY="..."
```

Define your steps:

```python
class Feedback(BaseModel):
    area: str
    message: str


class FeedbackCorpus(BaseModel):
    records: list[Feedback]


@ava.source
def load_corpus() -> FeedbackCorpus:
    return warehouse.load_feedback()
```

```python
class Theme(BaseModel):
    name: str
    evidence: list[str]


class ThemeReport(BaseModel):
    themes: list[Theme]


class ExtractThemes(ava.Signature):
    """Find recurring product needs and cite the feedback behind each one."""

    corpus: FeedbackCorpus = ava.InputField(
        desc="The complete product-feedback corpus to analyze."
    )
    report: ThemeReport = ava.OutputField(
        desc="Recurring themes with the supporting feedback."
    )


@ava.agent_step(ExtractThemes, lm="openai/gpt-5.5")
async def extract_themes(
    corpus: FeedbackCorpus,
    *,
    agent: ava.Agent,
) -> ThemeReport:
    return (await agent(corpus=corpus)).report
```

```python
class Risk(BaseModel):
    issue: str
    severity: Literal["low", "medium", "high"]
    evidence: list[str]


class RiskReport(BaseModel):
    risks: list[Risk]


class DetectRisks(ava.Signature):
    """Find product or customer risks and grade their severity."""

    corpus: FeedbackCorpus = ava.InputField(
        desc="The complete product-feedback corpus to analyze."
    )
    report: RiskReport = ava.OutputField(
        desc="Risks, severity, and the supporting feedback."
    )


@ava.agent_step(DetectRisks, lm="openai/gpt-5.5")
async def detect_risks(
    corpus: FeedbackCorpus,
    *,
    agent: ava.Agent,
) -> RiskReport:
    return (await agent(corpus=corpus)).report
```

```python
class CrmProductSignal(BaseModel):
    kind: Literal["risk", "theme"]
    headline: str
    evidence: list[str]


class CrmProductSignalBatch(BaseModel):
    signals: list[CrmProductSignal]


@ava.step
def compose_crm_product_signals(
    themes: ThemeReport,
    risks: RiskReport,
) -> CrmProductSignalBatch:
    return CrmProductSignalBatch(
        signals=[
            CrmProductSignal(
                kind="risk",
                headline=risk.issue,
                evidence=risk.evidence,
            )
            for risk in risks.risks
        ]
        + [
            CrmProductSignal(
                kind="theme",
                headline=theme.name,
                evidence=theme.evidence,
            )
            for theme in themes.themes
        ]
    )
```

```python
class CrmTask(BaseModel):
    external_id: str
    title: str


class CrmSyncResult(BaseModel):
    created: list[CrmTask]


@ava.dest
def push_to_crm(signals: CrmProductSignalBatch) -> CrmSyncResult:
    return crm.create_product_signals(signals)
```

Then chain them in a workflow:

```python
@ava.workflow
def feedback_triage():
    return (
        load_corpus()
        >> (extract_themes() & detect_risks())
        >> compose_crm_product_signals()
        >> push_to_crm()
    )
```

Run it through the local operator:

```bash
uv run ava dev --flows path/to/flow
```

The resulting graph has four execution stages:

```text
                         ┌─ extract_themes (agent) ─┐
load_corpus (source) ────┤                          ├─ compose_crm_product_signals ──→ push_to_crm
                         └─ detect_risks (agent) ───┘
```

Both analysis agents receive the same `FeedbackCorpus`. Avalanche runs them
concurrently, then binds their `ThemeReport` and `RiskReport` outputs into the
deterministic `compose_crm_product_signals` step in branch order. It maps the
reports into the CRM request shape. `push_to_crm` is a typed destination standing
in for a real service write and returns the external records it created.

Change the `lm` value and provider credentials to use another model supported by  
PredictRLM/DSPy. See [Agent steps](docs/agent-steps.md) for skills, tools,  
multi-output signatures, runtime defaults, and larger typed contracts.

## Terminal UI

Avalanche includes a Textual TUI for observing and controlling operator-managed
workflows from the terminal.

The TUI provides:

- workflow and run navigation;
- live DAG status;
- run history and node-level details;
- searchable logs;
- start, cancel, and rerun controls;
- schedule visibility and control.

Install the local control plane and TUI:

```bash
uv add "avalanche-ai[runtime,tui]"
```

The operator listens on `127.0.0.1` by default because its gRPC service does not
provide built-in authentication. Binding another interface with `--host` is an
explicit deployment choice and requires an external trusted and authenticated
boundary. Loopback limits network reachability; it does not identify callers,
and any process running as a local user may attempt to call the service.

In another terminal, start a run with JSON fields and a top-level file input.
`--file FIELD=PATH` reads and attaches the file bytes; it does not send the local
path to the operator:

```bash
RUN_ID=$(
  uv run ava run document_file_workflow \
    --connect localhost:7433 \
    --input '{"value": 41}' \
    --file document=./doc.txt
)
```

This command targets the bundled
[`document_file_workflow`](examples/document_file_workflow.py) example.

Download the successful terminal result without printing binary bytes. `--wait`
waits up to `--timeout` seconds for a nonterminal run:

```bash
uv run ava result "$RUN_ID" \
  --connect localhost:7433 \
  --wait \
  --output-dir ./run-result
```

The output directory must not already exist, and its parent directory must
exist. The CLI writes and verifies the complete result in a private staging
directory inside a retained, identity-pinned holding directory, syncs it, and
atomically renames the staged name without replacing an existing destination.
It immediately opens the requested destination through the retained parent
descriptor and compares its identity to the retained staging descriptor.
Substitution fails closed and triggers bounded, descriptor-anchored cleanup.
The published directory contains collision-resistant attachment filenames and a
generated `result-<uuid>.json` with the run ID, reconstructed result shape,
original file metadata, relative paths, sizes, and SHA-256 digests. See
[Run input and context](docs/dag-api.md#run-input-and-context) and
[Workflow results](docs/dag-api.md#workflow-results) for the Python and CLI
contracts, limits, and complete output layout.

The output parent is a caller-owned local namespace. POSIX and macOS provide no
portable operation that renames an open directory descriptor, or conditionally
renames a source name only if it still identifies a specific inode. A hostile
concurrent process running as the same user can therefore create a transient
wrong destination before the CLI detects and removes it; that concurrency is
outside this local CLI threat model. Descriptor-authenticated catchable state is
cleaned, and catchable failures leave no requested destination. An interruption
after the holding `mkdir` side effect but before descriptor acquisition can
leave private empty holding residue: safe cleanup cannot distinguish the
created directory from a same-name replacement, so it does not open, adopt, or
remove that entry. The requested destination remains absent. An uncatchable
termination such as `SIGKILL` can also leave private holding residue.

Or connect the TUI and start runs interactively:
Start the operator and connected TUI together:

```bash
ava dev --flows path/to/flows
```

Or run them separately:

```bash
# Terminal 1
ava operator --flows path/to/flows --port 7433

# Terminal 2
ava tui --connect localhost:7433
```

The TUI is a client of the operator; it does not import or execute workflow files
itself. To explore the interface without an operator, start mock mode:

```bash
python -m avalanche.tui
```



## Optional components


| Extra     | Purpose                                       |
| --------- | --------------------------------------------- |
| `agent`   | PredictRLM-backed agent steps                 |
| `runtime` | Operator, gRPC, file watching, and scheduling |
| `tui`     | Textual terminal UI                           |
| `ray`     | Ray-backed workflow execution                 |
| `lance`   | Lance storage backend                         |
| `all`     | Every optional component above                |


Extras can be combined:

```bash
uv add "avalanche-ai[agent,runtime,tui]"
```



## Documentation

- [Getting started](docs/getting-started.md)
- [DAG API](docs/dag-api.md)
- [Agent steps](docs/agent-steps.md)
- [Data model and storage API](docs/data-model-api.md)
- [Execution services](docs/execution-services.md)
- [Architecture](ARCHITECTURE.md)
- [Examples](examples/README.md)
- [Changelog](CHANGELOG.md)



## Project status

Avalanche currently targets local development and experimentation. Production
authentication, authorization, TLS, multitenancy, one-click deployment, and full
durable operator recovery are not implemented. Point operator discovery at a
specific workflow file or clean workflow directory rather than a repository root,
because discovery imports Python modules.

## Contributing

Contributions are welcome. See [CONTRIBUTING.md](CONTRIBUTING.md) for local setup,
quality gates, and pull request expectations.

## License

Avalanche is licensed under the [Apache License 2.0](LICENSE).