reflex-agentkit 0.1.0__tar.gz

reflex_agentkit-0.1.0/.gitignore

@@ -0,0 +1,229 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#   Usually these files are written by a python script from a template
+#   before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+*.lcov
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+# Pipfile.lock
+
+# uv
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+# uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+# poetry.lock
+# poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+# pdm.lock
+# pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+# pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi/*
+!.pixi/config.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule*
+celerybeat.pid
+
+# Redis
+*.rdb
+*.aof
+*.pid
+
+# RabbitMQ
+mnesia/
+rabbitmq/
+rabbitmq-data/
+
+# ActiveMQ
+activemq-data/
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs/Zensical documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#   JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#   be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#   and can be added to the global gitignore or merged into this file.  For a more nuclear
+#   option (not recommended) you can uncomment the following to ignore the entire idea folder.
+# .idea/
+
+# Abstra
+#   Abstra is an AI-powered process automation framework.
+#   Ignore directories containing user credentials, local state, and settings.
+#   Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#   Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore that
+#   can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#   and can be added to the global gitignore or merged into this file. However, if you prefer, you
+#   could uncomment the following to ignore the entire vscode folder
+# .vscode/
+# Temporary file for partial code execution
+tempCodeRunnerFile.py
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+
+# Streamlit
+.streamlit/secrets.toml
+
+# JetBrains
+.idea/
+
+# reflex-agentkit
+.venv/
+.web/
+dist/
+.coverage

reflex_agentkit-0.1.0/CHANGELOG.md

@@ -0,0 +1,121 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+## [0.1.0] - 2026-10-31
+
+The first release. `0.1.0rc1` was published on 2026-10-31 as its release candidate.
+
+### Added
+
+- Project skeleton: uv/hatchling packaging, lint, type and test configuration, and CI.
+- `reflex_agentkit.core`: error hierarchy and `ErrorCode`, the `RunStatus` and
+  `ToolCallStatus` enums, the value types (`ChatMessage`, `ToolCall`, `PendingInterrupt`,
+  `CustomEventRecord`, `Diagnostic`) and the `ConversationState` aggregate, all with
+  version-tolerant pickling.
+- `reflex_agentkit.core.reducer`: the pure, total `reduce(state, event)` over all 31
+  AG-UI 1.0 event types (lifecycle and `RUN_FINISHED` outcomes, text, tool calls,
+  shared state, messages snapshots, reasoning, custom events) and `parse_event`, with
+  diagnostics instead of exceptions.
+- `reflex_agentkit.core.commands`: `begin_run`, `apply_approval` (with edited arguments),
+  `apply_rejection`, `apply_interrupt_response`, `apply_cancel`,
+  `apply_client_tool_result`, `all_answered`, `resume_entries` and `new_thread`.
+- `client_tool` / `ClientTool`: client-side tools with a JSON Schema built from the type
+  hints, `to_agui_tool()` and `invoke()` (sync handlers in a worker thread, timeout,
+  failures returned as tool results).
+- `RunInputBuilder`: builds `RunAgentInput` from the conversation (full history,
+  encrypted reasoning before its assistant message, edited tool arguments, shared state,
+  client tools, resume entries).
+- Agent clients: the runtime-checkable `AGUIAgent` / `AgentClient` protocols and
+  `AgentSource`; `HttpAGUIClient` (AG-UI over HTTP + SSE, static or per-run headers that
+  are never logged, read timeout, typed `AgentHTTPError` / `AgentStreamError`, upstream
+  closed on cancel) and `InProcessAGUIClient` (runs an agent object in the event loop,
+  cancellation reaches the agent); an incremental SSE parser; invalid events are skipped
+  with a log record.
+- `AgentState`, the Reflex state mixin: `ClassVar` configuration validated when the class
+  is defined (`ConfigurationError` names the attribute); public vars projected from a
+  backend-only conversation (system messages, interrupt answers and encrypted reasoning
+  never reach the browser); computed vars `is_running`, `awaiting_input`,
+  `pending_approvals` and `can_send`; handlers `send_message`, `approve_tool`,
+  `reject_tool`, `respond_to_interrupt`, `cancel_run`, `new_thread`, `set_shared_state`,
+  `patch_shared_state` and `retry_run`; hooks `create_client`, `prepare_run_input`,
+  `on_custom_event` and `on_run_end`; `typed_shared_state()` with `shared_state_model`
+  validation of app writes and agent updates.
+- The run driver behind `AgentState`: streaming in a background task with batched pushes
+  (default 50 ms), cancellation within 1 s in the same worker or from another worker over
+  a shared state manager, client-tool rounds with follow-up runs, local approvals and a
+  loop limit, and silent stop when the conversation was replaced.
+- `reflex_agentkit.adapters.pydantic_ai.PydanticAIAgent`: runs a PydanticAI agent in
+  process through its AG-UI adapter; approval-required tools pause as AG-UI interrupts
+  without declaring `DeferredToolRequests`; `deps` or a per-run `deps_factory`; extra
+  `run_stream` arguments are forwarded. Importing an adapter without its extra raises
+  `AdapterNotInstalledError` with the install command.
+- `scripts/record_fixture.py` records event fixtures from real adapter runs.
+- `reflex_agentkit.components`: `agent_chat` and its parts, built from Reflex built-ins
+  only. `message_list` renders Markdown with raw HTML off, reasoning and a streaming
+  indicator; `tool_call_card` shows a status badge and collapsible arguments and result;
+  `composer` sends on Enter; `run_status_bar` has a cancel button; `approval_dialog`
+  offers approve, edit-arguments and reject, or a form for generic interrupts with their
+  queue position; `shared_state_panel` shows the JSON and a stale marker. `ToolRenderer`
+  and `InterruptRenderer` replace the default card body or form. Every control has an
+  accessible label.
+- `agent_chat` is re-exported lazily from `reflex_agentkit`, and `reduce` from
+  `reflex_agentkit.core`, as in API §3.
+- `examples/quickstart/`: the 23-line quickstart, a uv workspace member, with an offline
+  CI smoke job (`reflex compile --dry` with PydanticAI's `TestModel`).
+- `examples/_shared/` (`examples-shared`): one model selection for every example —
+  Claude by default, `AGENTKIT_MODEL=ollama:<model>` for a local Ollama model, and
+  `AGENTKIT_MODEL=test` for an offline model — for PydanticAI and for LangChain chat
+  models.
+- `examples/pydantic_ai_dojo/` and `examples/langgraph_dojo/`: the five AG-UI Dojo demos
+  (agentic chat, backend tool rendering, human in the loop, shared state, tool-based
+  generative UI) as multi-page Reflex apps, with offline smoke tests and CI compile steps.
+- Automatic releases: merging a new publishable `__version__` into `main` runs the tests,
+  publishes to PyPI through trusted publishing and creates the `v<version>` tag and GitHub
+  Release (`scripts/release_version.py` decides). Pre-releases go to PyPI too.
+- `reflex_agentkit.adapters.langgraph`: `LangGraphAGUIAgent` runs a compiled LangGraph
+  graph through the official ag-ui-langgraph integration with standard interrupts. It
+  requires a checkpointer, serves every session from one instance, keeps the graph's
+  messages out of `shared_state`, and resumes several interrupts with LangGraph's native
+  map. `request_tool_approval` and `ApprovalDecision` put tool approvals in the same
+  dialog as PydanticAI's. `legacy_interrupts=True` maps the older `on_interrupt` event.
+- A draft upstream issue for ag-ui-langgraph's unconditional `fastapi` import
+  (`docs/integrations/upstream-issues.md`).
+- JSON Lines event fixtures and a replay runner that checks the state invariants after
+  every step.
+- A documentation site (MkDocs Material, `mkdocs.yml` and `docs/`): getting started,
+  concepts, guides (your own UI, human in the loop, sidecar deployment, Ollama), the API
+  reference, the security model and troubleshooting. CI builds it with `--strict`.
+- Benchmarks under the `bench` marker: the reducer (NFR-01) and the flush lag with 20
+  concurrent clients (NFR-04).
+- A weekly workflow that tests against the newest allowed dependency versions and opens
+  an issue when they fail.
+
+### Fixed
+
+- A tool call that a model sends in one chunk (Ollama through LangGraph) now shows its
+  arguments in the approval dialog: a messages snapshot fills in the arguments of a call
+  that streamed none. Found in the `0.1.0rc1` walkthrough.
+- A `TOOL_CALL_RESULT` no longer moves a rejected or failed tool call back to `done`.
+- A `MESSAGES_SNAPSHOT` now fills in the result of a tool call it keeps, so LangGraph tool
+  calls end as `done` after their tool node runs.
+
+### Security
+
+- Reducer diagnostics are logged above DEBUG by code and event type only, since a failed
+  patch can quote shared-state values.
+- App-side shared-state patches are applied op by op under the 64 KiB cap and limited to
+  256 operations, so `copy` ops cannot grow the document exponentially.
+- `AgentState`'s hooks (`create_client`, `prepare_run_input`, `on_custom_event`,
+  `on_run_end`) and `typed_shared_state`, and every override of them, are no longer
+  registered as Reflex event handlers, so the browser cannot call them.
+- The latest-dependencies workflow runs the newest, unvetted versions without a token or
+  stored git credentials; a separate job reports failures.
+
+[Unreleased]: https://github.com/ecrespo/reflex-agentkit/compare/v0.1.0...HEAD
+[0.1.0]: https://github.com/ecrespo/reflex-agentkit/releases/tag/v0.1.0

reflex_agentkit-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Ernesto Crespo
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.