drift-detection 0.1.0__tar.gz

drift_detection-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,21 @@
+name: Publish to PyPI
+
+on:
+  push:
+    tags:
+      - "v*"
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      id-token: write
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - run: pip install build
+      - run: python -m build
+      - uses: pypa/gh-action-pypi-publish@release/v1

drift_detection-0.1.0/.gitignore

@@ -0,0 +1,218 @@
+# 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
+.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
+
+# 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 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

drift_detection-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 dombinic
+
+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.

drift_detection-0.1.0/PKG-INFO

@@ -0,0 +1,181 @@
+Metadata-Version: 2.4
+Name: drift-detection
+Version: 0.1.0
+Summary: Statistical anomaly detection for AI agent workflows
+Project-URL: Homepage, https://github.com/dombinic/Drift
+Project-URL: Repository, https://github.com/dombinic/Drift
+Project-URL: Issues, https://github.com/dombinic/Drift/issues
+Author: Dominic
+License: MIT
+License-File: LICENSE
+Keywords: agents,ai,anomaly-detection,langchain,llm,monitoring,observability
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Requires-Python: >=3.9
+Requires-Dist: numpy>=1.24.0
+Provides-Extra: all
+Requires-Dist: langchain-core>=0.1.0; extra == 'all'
+Provides-Extra: dev
+Requires-Dist: pytest-asyncio>=0.21; extra == 'dev'
+Requires-Dist: pytest>=7.0; extra == 'dev'
+Requires-Dist: ruff>=0.1.0; extra == 'dev'
+Provides-Extra: langchain
+Requires-Dist: langchain-core>=0.1.0; extra == 'langchain'
+Description-Content-Type: text/markdown
+
+# Drift
+
+**Statistical anomaly detection for AI agent workflows.**
+
+Catch silent failures, hallucination drift, and off-script behavior before they corrupt your data.
+
+---
+
+Your AI agents are failing silently. A tool call takes 10x longer than usual. The agent calls `delete_file` when it's never done that before. Output quality gradually degrades over hundreds of runs. Traditional monitoring tools weren't built for non-deterministic systems — Drift is.
+
+## What it does
+
+Drift hooks into your agent's execution and applies statistical anomaly detection to the event stream:
+
+- **Latency & token SPC** — Flags when a tool call or LLM response takes significantly longer or uses significantly more tokens than its rolling baseline. Catches hung API calls, runaway generation, and upstream provider issues.
+
+- **Sequence anomaly detection** — Builds a transition matrix of tool-call sequences and flags when the agent takes a path that's never or rarely been seen. Catches agents going off-script, skipping required steps, or entering novel execution paths.
+
+- **Output drift detection** — Tracks output length, vocabulary diversity, and structural patterns over time. Flags when outputs shift significantly from baseline. Catches hallucination drift, prompt injection effects, and gradual quality degradation.
+
+## Install
+
+```bash
+pip install driftguard
+```
+
+With LangChain support:
+```bash
+pip install driftguard[langchain]
+```
+
+## Quickstart
+
+### With LangChain
+
+```python
+from drift import DriftGuard
+from drift.callbacks.langchain import DriftCallbackHandler
+
+guard = DriftGuard(on_anomaly=lambda a: print(f"🚨 {a}"))
+handler = DriftCallbackHandler(guard)
+
+# Use with any LangChain agent, chain, or LLM
+agent.run("your query", callbacks=[handler])
+
+# See what happened
+guard.report()
+```
+
+### Standalone (no framework required)
+
+```python
+from drift import DriftGuard, AgentEvent, EventType
+
+guard = DriftGuard(on_anomaly=lambda a: print(f"🚨 {a}"))
+
+# Feed events from any source
+guard.ingest(AgentEvent(
+    event_type=EventType.TOOL_END,
+    name="search_web",
+    latency_ms=150.0,
+    token_count=85,
+    output_text="Found 3 results for query...",
+))
+
+guard.report()
+```
+
+### Run the demo
+
+```bash
+python examples/demo.py
+```
+
+This simulates normal agent operation, builds baselines, then injects latency spikes, sequence anomalies, and output drift — showing each detector catching real failure modes.
+
+## Architecture
+
+```
+drift/
+├── core.py              # DriftGuard engine — orchestrates detectors
+├── models.py            # AgentEvent, Anomaly, Severity data models
+├── detectors/
+│   ├── latency.py       # Statistical process control on latency/tokens
+│   ├── sequence.py      # Action transition probability anomalies
+│   └── output_drift.py  # Output distribution shift detection
+└── callbacks/
+    └── langchain.py     # LangChain callback integration
+```
+
+**Design principles:**
+
+1. **Zero-overhead default** — Detectors use numpy for fast rolling statistics. No embedding models, no external services, no network calls.
+2. **Per-tool baselines** — Each tool and model gets its own statistical baseline, so a slow tool won't pollute the baseline for a fast one.
+3. **Framework-agnostic core** — The detection engine works with raw `AgentEvent` objects. Framework integrations (LangChain, CrewAI, etc.) are thin adapters that translate framework callbacks into events.
+4. **Non-blocking** — Drift never throws exceptions that would crash your agent. Detector errors are caught and logged to stderr.
+
+## Detectors
+
+| Detector | What it catches | Method |
+|----------|----------------|--------|
+| `LatencyDetector` | Hung calls, slow APIs, runaway generation | Rolling z-score on latency and token counts |
+| `SequenceDetector` | Off-script behavior, unexpected tool calls | First-order Markov transition probabilities |
+| `OutputDriftDetector` | Hallucination drift, prompt injection, quality degradation | Output length, vocab diversity, structural pattern tracking |
+
+## Configuration
+
+Each detector is independently configurable:
+
+```python
+from drift import DriftGuard
+from drift.detectors.latency import LatencyDetector, LatencyDetectorConfig
+from drift.detectors.sequence import SequenceDetector, SequenceDetectorConfig
+
+guard = DriftGuard(detectors=[
+    LatencyDetector(LatencyDetectorConfig(
+        window_size=100,     # Longer baseline window
+        z_threshold=2.5,     # More sensitive
+        min_samples=10,      # Require more data before alerting
+    )),
+    SequenceDetector(SequenceDetectorConfig(
+        min_observations=20, # Require more transitions before flagging
+    )),
+])
+```
+
+## Roadmap
+
+- [ ] CrewAI callback handler
+- [ ] OpenAI Agents SDK integration
+- [ ] Slack / PagerDuty alerting
+- [ ] Persistent baselines (save/load detector state)
+- [ ] Embedding-based output drift (optional dependency)
+- [ ] Web dashboard
+- [ ] Cost anomaly detection (track spend per run)
+
+## Contributing
+
+Issues and PRs welcome. Run tests with:
+
+```bash
+pip install -e ".[dev]"
+pytest
+```
+
+## License
+
+MIT

drift_detection-0.1.0/README.md

@@ -0,0 +1,148 @@
+# Drift
+
+**Statistical anomaly detection for AI agent workflows.**
+
+Catch silent failures, hallucination drift, and off-script behavior before they corrupt your data.
+
+---
+
+Your AI agents are failing silently. A tool call takes 10x longer than usual. The agent calls `delete_file` when it's never done that before. Output quality gradually degrades over hundreds of runs. Traditional monitoring tools weren't built for non-deterministic systems — Drift is.
+
+## What it does
+
+Drift hooks into your agent's execution and applies statistical anomaly detection to the event stream:
+
+- **Latency & token SPC** — Flags when a tool call or LLM response takes significantly longer or uses significantly more tokens than its rolling baseline. Catches hung API calls, runaway generation, and upstream provider issues.
+
+- **Sequence anomaly detection** — Builds a transition matrix of tool-call sequences and flags when the agent takes a path that's never or rarely been seen. Catches agents going off-script, skipping required steps, or entering novel execution paths.
+
+- **Output drift detection** — Tracks output length, vocabulary diversity, and structural patterns over time. Flags when outputs shift significantly from baseline. Catches hallucination drift, prompt injection effects, and gradual quality degradation.
+
+## Install
+
+```bash
+pip install driftguard
+```
+
+With LangChain support:
+```bash
+pip install driftguard[langchain]
+```
+
+## Quickstart
+
+### With LangChain
+
+```python
+from drift import DriftGuard
+from drift.callbacks.langchain import DriftCallbackHandler
+
+guard = DriftGuard(on_anomaly=lambda a: print(f"🚨 {a}"))
+handler = DriftCallbackHandler(guard)
+
+# Use with any LangChain agent, chain, or LLM
+agent.run("your query", callbacks=[handler])
+
+# See what happened
+guard.report()
+```
+
+### Standalone (no framework required)
+
+```python
+from drift import DriftGuard, AgentEvent, EventType
+
+guard = DriftGuard(on_anomaly=lambda a: print(f"🚨 {a}"))
+
+# Feed events from any source
+guard.ingest(AgentEvent(
+    event_type=EventType.TOOL_END,
+    name="search_web",
+    latency_ms=150.0,
+    token_count=85,
+    output_text="Found 3 results for query...",
+))
+
+guard.report()
+```
+
+### Run the demo
+
+```bash
+python examples/demo.py
+```
+
+This simulates normal agent operation, builds baselines, then injects latency spikes, sequence anomalies, and output drift — showing each detector catching real failure modes.
+
+## Architecture
+
+```
+drift/
+├── core.py              # DriftGuard engine — orchestrates detectors
+├── models.py            # AgentEvent, Anomaly, Severity data models
+├── detectors/
+│   ├── latency.py       # Statistical process control on latency/tokens
+│   ├── sequence.py      # Action transition probability anomalies
+│   └── output_drift.py  # Output distribution shift detection
+└── callbacks/
+    └── langchain.py     # LangChain callback integration
+```
+
+**Design principles:**
+
+1. **Zero-overhead default** — Detectors use numpy for fast rolling statistics. No embedding models, no external services, no network calls.
+2. **Per-tool baselines** — Each tool and model gets its own statistical baseline, so a slow tool won't pollute the baseline for a fast one.
+3. **Framework-agnostic core** — The detection engine works with raw `AgentEvent` objects. Framework integrations (LangChain, CrewAI, etc.) are thin adapters that translate framework callbacks into events.
+4. **Non-blocking** — Drift never throws exceptions that would crash your agent. Detector errors are caught and logged to stderr.
+
+## Detectors
+
+| Detector | What it catches | Method |
+|----------|----------------|--------|
+| `LatencyDetector` | Hung calls, slow APIs, runaway generation | Rolling z-score on latency and token counts |
+| `SequenceDetector` | Off-script behavior, unexpected tool calls | First-order Markov transition probabilities |
+| `OutputDriftDetector` | Hallucination drift, prompt injection, quality degradation | Output length, vocab diversity, structural pattern tracking |
+
+## Configuration
+
+Each detector is independently configurable:
+
+```python
+from drift import DriftGuard
+from drift.detectors.latency import LatencyDetector, LatencyDetectorConfig
+from drift.detectors.sequence import SequenceDetector, SequenceDetectorConfig
+
+guard = DriftGuard(detectors=[
+    LatencyDetector(LatencyDetectorConfig(
+        window_size=100,     # Longer baseline window
+        z_threshold=2.5,     # More sensitive
+        min_samples=10,      # Require more data before alerting
+    )),
+    SequenceDetector(SequenceDetectorConfig(
+        min_observations=20, # Require more transitions before flagging
+    )),
+])
+```
+
+## Roadmap
+
+- [ ] CrewAI callback handler
+- [ ] OpenAI Agents SDK integration
+- [ ] Slack / PagerDuty alerting
+- [ ] Persistent baselines (save/load detector state)
+- [ ] Embedding-based output drift (optional dependency)
+- [ ] Web dashboard
+- [ ] Cost anomaly detection (track spend per run)
+
+## Contributing
+
+Issues and PRs welcome. Run tests with:
+
+```bash
+pip install -e ".[dev]"
+pytest
+```
+
+## License
+
+MIT

drift_detection-0.1.0/drift/__init__.py

@@ -0,0 +1,28 @@
+"""Drift — Statistical anomaly detection for AI agent workflows.
+
+Catch silent failures, hallucination drift, and off-script behavior
+in your LangChain, CrewAI, and custom AI agents.
+
+Quickstart:
+    from drift import DriftGuard
+    from drift.callbacks.langchain import DriftCallbackHandler
+
+    guard = DriftGuard()
+    handler = DriftCallbackHandler(guard)
+    agent.run("your query", callbacks=[handler])
+    guard.report()
+"""
+
+from drift.core import DriftGuard
+from drift.models import AgentEvent, Anomaly, AnomalyType, EventType, Severity
+
+__version__ = "0.1.0"
+
+__all__ = [
+    "DriftGuard",
+    "AgentEvent",
+    "Anomaly",
+    "AnomalyType",
+    "EventType",
+    "Severity",
+]

drift_detection-0.1.0/drift/callbacks/__init__.py

@@ -0,0 +1 @@
+"""Framework callback integrations for Drift."""