tracelint 0.1.0__tar.gz

tracelint-0.1.0/.gitattributes

@@ -0,0 +1,5 @@
+# Normalize line endings: LF in the repo, checked out native on Windows.
+* text=auto eol=lf
+*.png binary
+*.jpg binary
+*.gif binary

tracelint-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,30 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10", "3.12"]
+    steps:
+      - uses: actions/checkout@v4
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Install
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+      - name: Ruff
+        run: ruff check src tests
+      - name: Pytest
+        run: python -m pytest -q
+      - name: Demo self-check (validation suite + scorecard, keyless)
+        run: tracelint demo

tracelint-0.1.0/.github/workflows/pages.yml

@@ -0,0 +1,53 @@
+name: Deploy demo report to Pages
+
+on:
+  push:
+    branches: [main]
+    paths:
+      - "src/**"
+      - ".github/workflows/pages.yml"
+  workflow_dispatch:
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+concurrency:
+  group: pages
+  cancel-in-progress: true
+
+jobs:
+  deploy:
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install tracelint
+        run: pip install -e .
+
+      - name: Generate the demo report
+        run: |
+          mkdir site
+          tracelint demo --html site/index.html
+
+      - name: Configure GitHub Pages
+        uses: actions/configure-pages@v5
+
+      - name: Upload Pages artifact
+        uses: actions/upload-pages-artifact@v3
+        with:
+          path: site
+
+      - name: Deploy to GitHub Pages
+        id: deployment
+        uses: actions/deploy-pages@v4

tracelint-0.1.0/.github/workflows/release.yml

@@ -0,0 +1,47 @@
+name: Publish to PyPI
+
+# Publishes on a GitHub Release using PyPI Trusted Publishing (OIDC) — no API token is
+# stored anywhere. One-time setup: add a Trusted Publisher on PyPI for project "tracelint"
+# (owner AshwinUgale, repo tracelint, workflow "release.yml", environment "pypi"), then
+# create a GitHub Release (tag v0.1.0) to trigger this workflow.
+
+on:
+  release:
+    types: [published]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - name: Build sdist and wheel
+        run: |
+          python -m pip install --upgrade build
+          python -m build
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+  publish:
+    needs: build
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/tracelint
+    permissions:
+      id-token: write  # required for Trusted Publishing (OIDC)
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

tracelint-0.1.0/.gitignore

@@ -0,0 +1,23 @@
+# Python
+__pycache__/
+*.py[cod]
+*.egg-info/
+.eggs/
+build/
+dist/
+.venv/
+venv/
+env/
+
+# Tooling
+.pytest_cache/
+.ruff_cache/
+.mypy_cache/
+.coverage
+htmlcov/
+
+# Editors / OS
+.vscode/
+.idea/
+.DS_Store
+Thumbs.db

tracelint-0.1.0/LICENSE

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

tracelint-0.1.0/PKG-INFO

@@ -0,0 +1,181 @@
+Metadata-Version: 2.4
+Name: tracelint
+Version: 0.1.0
+Summary: Deterministic, judge-free static analyzer for tool-calling agent execution traces.
+Project-URL: Homepage, https://github.com/AshwinUgale/tracelint
+Project-URL: Repository, https://github.com/AshwinUgale/tracelint
+Project-URL: Issues, https://github.com/AshwinUgale/tracelint/issues
+Author: Ashwin Ugale
+License: MIT License
+        
+        Copyright (c) 2026 Ashwin Ugale
+        
+        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.
+License-File: LICENSE
+Keywords: agents,linter,llm,observability,reliability,static-analysis,tool-calling,traces
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Quality Assurance
+Classifier: Topic :: Software Development :: Testing
+Requires-Python: >=3.10
+Requires-Dist: jsonschema>=4.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7; extra == 'dev'
+Requires-Dist: ruff>=0.4; extra == 'dev'
+Provides-Extra: real-agent
+Requires-Dist: openai>=1.0; extra == 'real-agent'
+Description-Content-Type: text/markdown
+
+# tracelint
+
+**ESLint/pytest for what your agent actually did.** `tracelint` is a deterministic, judge-free
+static analyzer for the execution traces of tool-calling agents. It reads a trace and reports
+structural defects — schema-violating tool calls, ignored tool errors, hallucinated arguments,
+loops, and redundant calls — each with the exact trace lines as evidence, and returns a CI exit
+code. It also ships a fault injector and a per-fault recovery scorecard.
+
+Model-as-judge detection of these defects is unreliable (published trace-error benchmarks show low
+localization accuracy). Many of these defects are *structurally decidable* and need no judge — that
+is the entire premise of this tool. No second model ever judges the trace.
+
+**[View the live demo report](https://ashwinugale.github.io/tracelint/)** — the constructed
+validation suite (one planted instance of every defect, clean controls, and legitimate-but-suspicious
+cases) plus the robust-vs-buggy recovery scorecard, generated by `tracelint demo`.
+
+## Limitations (read first)
+
+1. Deterministic rules catch **structural** defects, not whether the final answer was correct.
+2. Hallucinated-argument, loop, and redundant-call findings are **candidates** unless structurally
+   proven — legitimate value transforms and intentional retries can trip them; each is shown with
+   its evidence for human review, never asserted as a verdict. High-confidence hallucination
+   detection requires the tool schema to declare field origins (`x-value-origin`).
+3. The recovery scorecard needs labeled task outcomes (success oracles); without them it measures
+   **behavioral** recovery only ("did not crash"), a weaker claim than correctness.
+4. A trace is only as complete as its instrumentation. A rule whose required field is missing is
+   **suppressed with a stated reason** — `tracelint` never lints a partial trace as if complete.
+
+## Quick start
+
+The demo runs a keyless validation suite and a recovery scorecard end to end — no API key, no
+model download:
+
+```bash
+pip install tracelint
+tracelint demo --html demo.html
+```
+
+Lint a trace in CI:
+
+```bash
+tracelint check ./trace.json --tools ./tools.json     # exit 2 on a hard_defect
+```
+
+Exit codes: `0` clean · `2` a structurally-provable defect (`hard_defect`) · `3` an input error.
+Heuristic candidates never fail CI on their own; suppressions are disclosed but are not defects.
+
+## The rules
+
+| Rule | Finding | Tiers |
+|------|---------|-------|
+| R1 | schema violation — args fail the tool's JSON Schema | `hard_defect` |
+| R2a | tool returned an error | `hard_event` (structured signal) / `candidate` (heuristic) |
+| R2b | an errored result's value reused by a later side-effecting call | `hard_defect` / `candidate` |
+| R3 | hallucinated argument — value not derivable from provenance | `candidate`; `hard_defect` if the field is annotated `provided` |
+| R4 | loop — N identical no-progress calls (polls/retries excluded) | `candidate` |
+| R5 | redundant call — identical call + identical result, no mutation between | `candidate` |
+
+`hard_event` and `hard_defect` are orthogonal to the finding kind: a tool-error event is a
+`hard_event` from a structured status field but a `candidate` from an exception-like string in
+free-form content.
+
+## Input format
+
+A trace is a JSON object (`.json`, or `.jsonl` for many):
+
+```json
+{
+  "run_id": "run-1",
+  "steps": [
+    {"type": "message", "role": "user", "content": "cancel order 4521 if it hasn't shipped"},
+    {"type": "tool_call", "call_id": "c1", "name": "get_order_status", "args": {"order_id": "4521"}},
+    {"type": "tool_result", "call_id": "c1", "content": {"status": "processing"}, "status": "ok"},
+    {"type": "tool_call", "call_id": "c2", "name": "cancel_order",
+     "args": {"order_id": "4521", "reason": "not_shipped"}}
+  ],
+  "final": "Order 4521 has been cancelled."
+}
+```
+
+`tools.json` supplies the ground truth the rules check against:
+
+```json
+{
+  "tools": {
+    "cancel_order": {
+      "schema": {"type": "object", "properties": {"order_id": {"type": "string"}},
+                 "required": ["order_id"]},
+      "metadata": {"side_effecting": true}
+    }
+  }
+}
+```
+
+An OpenAI adapter (`tracelint.adapters.from_openai_messages`) normalizes OpenAI chat message lists
+into this schema; more adapters are future work.
+
+## Recovery scorecard
+
+Measure how an agent behaves under injected faults, scored against deterministic success oracles:
+
+```bash
+tracelint scorecard --demo --faults timeout,error,rate_limit --runs 5
+```
+
+The baseline must satisfy the oracle first (else recovery is not measured). Each fault type reports
+a correctness-recovery rate with a Wilson confidence interval; with no oracle it falls back to
+behavioral recovery, labeled as weaker.
+
+## Library
+
+```python
+from tracelint import lint_trace, default_rules, Trace, ToolRegistry
+
+trace = Trace.load("trace.json")
+registry = ToolRegistry.load("tools.json")
+report = lint_trace(trace, default_rules(), registry)
+print(report.exit_code)          # 0 or 2
+for f in report.active_findings:
+    print(f.rule, f.tier.value, f.summary)
+```
+
+## Development
+
+```bash
+python -m pytest
+ruff check src tests
+```
+
+The core is dependency-light (`jsonschema` + stdlib) and the whole test suite is deterministic and
+offline. A real OpenAI trace-generating agent lives behind the opt-in `[real-agent]` extra and is
+never part of the linter. Python 3.10–3.12.

tracelint-0.1.0/README.md

@@ -0,0 +1,132 @@
+# tracelint
+
+**ESLint/pytest for what your agent actually did.** `tracelint` is a deterministic, judge-free
+static analyzer for the execution traces of tool-calling agents. It reads a trace and reports
+structural defects — schema-violating tool calls, ignored tool errors, hallucinated arguments,
+loops, and redundant calls — each with the exact trace lines as evidence, and returns a CI exit
+code. It also ships a fault injector and a per-fault recovery scorecard.
+
+Model-as-judge detection of these defects is unreliable (published trace-error benchmarks show low
+localization accuracy). Many of these defects are *structurally decidable* and need no judge — that
+is the entire premise of this tool. No second model ever judges the trace.
+
+**[View the live demo report](https://ashwinugale.github.io/tracelint/)** — the constructed
+validation suite (one planted instance of every defect, clean controls, and legitimate-but-suspicious
+cases) plus the robust-vs-buggy recovery scorecard, generated by `tracelint demo`.
+
+## Limitations (read first)
+
+1. Deterministic rules catch **structural** defects, not whether the final answer was correct.
+2. Hallucinated-argument, loop, and redundant-call findings are **candidates** unless structurally
+   proven — legitimate value transforms and intentional retries can trip them; each is shown with
+   its evidence for human review, never asserted as a verdict. High-confidence hallucination
+   detection requires the tool schema to declare field origins (`x-value-origin`).
+3. The recovery scorecard needs labeled task outcomes (success oracles); without them it measures
+   **behavioral** recovery only ("did not crash"), a weaker claim than correctness.
+4. A trace is only as complete as its instrumentation. A rule whose required field is missing is
+   **suppressed with a stated reason** — `tracelint` never lints a partial trace as if complete.
+
+## Quick start
+
+The demo runs a keyless validation suite and a recovery scorecard end to end — no API key, no
+model download:
+
+```bash
+pip install tracelint
+tracelint demo --html demo.html
+```
+
+Lint a trace in CI:
+
+```bash
+tracelint check ./trace.json --tools ./tools.json     # exit 2 on a hard_defect
+```
+
+Exit codes: `0` clean · `2` a structurally-provable defect (`hard_defect`) · `3` an input error.
+Heuristic candidates never fail CI on their own; suppressions are disclosed but are not defects.
+
+## The rules
+
+| Rule | Finding | Tiers |
+|------|---------|-------|
+| R1 | schema violation — args fail the tool's JSON Schema | `hard_defect` |
+| R2a | tool returned an error | `hard_event` (structured signal) / `candidate` (heuristic) |
+| R2b | an errored result's value reused by a later side-effecting call | `hard_defect` / `candidate` |
+| R3 | hallucinated argument — value not derivable from provenance | `candidate`; `hard_defect` if the field is annotated `provided` |
+| R4 | loop — N identical no-progress calls (polls/retries excluded) | `candidate` |
+| R5 | redundant call — identical call + identical result, no mutation between | `candidate` |
+
+`hard_event` and `hard_defect` are orthogonal to the finding kind: a tool-error event is a
+`hard_event` from a structured status field but a `candidate` from an exception-like string in
+free-form content.
+
+## Input format
+
+A trace is a JSON object (`.json`, or `.jsonl` for many):
+
+```json
+{
+  "run_id": "run-1",
+  "steps": [
+    {"type": "message", "role": "user", "content": "cancel order 4521 if it hasn't shipped"},
+    {"type": "tool_call", "call_id": "c1", "name": "get_order_status", "args": {"order_id": "4521"}},
+    {"type": "tool_result", "call_id": "c1", "content": {"status": "processing"}, "status": "ok"},
+    {"type": "tool_call", "call_id": "c2", "name": "cancel_order",
+     "args": {"order_id": "4521", "reason": "not_shipped"}}
+  ],
+  "final": "Order 4521 has been cancelled."
+}
+```
+
+`tools.json` supplies the ground truth the rules check against:
+
+```json
+{
+  "tools": {
+    "cancel_order": {
+      "schema": {"type": "object", "properties": {"order_id": {"type": "string"}},
+                 "required": ["order_id"]},
+      "metadata": {"side_effecting": true}
+    }
+  }
+}
+```
+
+An OpenAI adapter (`tracelint.adapters.from_openai_messages`) normalizes OpenAI chat message lists
+into this schema; more adapters are future work.
+
+## Recovery scorecard
+
+Measure how an agent behaves under injected faults, scored against deterministic success oracles:
+
+```bash
+tracelint scorecard --demo --faults timeout,error,rate_limit --runs 5
+```
+
+The baseline must satisfy the oracle first (else recovery is not measured). Each fault type reports
+a correctness-recovery rate with a Wilson confidence interval; with no oracle it falls back to
+behavioral recovery, labeled as weaker.
+
+## Library
+
+```python
+from tracelint import lint_trace, default_rules, Trace, ToolRegistry
+
+trace = Trace.load("trace.json")
+registry = ToolRegistry.load("tools.json")
+report = lint_trace(trace, default_rules(), registry)
+print(report.exit_code)          # 0 or 2
+for f in report.active_findings:
+    print(f.rule, f.tier.value, f.summary)
+```
+
+## Development
+
+```bash
+python -m pytest
+ruff check src tests
+```
+
+The core is dependency-light (`jsonschema` + stdlib) and the whole test suite is deterministic and
+offline. A real OpenAI trace-generating agent lives behind the opt-in `[real-agent]` extra and is
+never part of the linter. Python 3.10–3.12.

tracelint-0.1.0/pyproject.toml

@@ -0,0 +1,66 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "tracelint"
+version = "0.1.0"
+description = "Deterministic, judge-free static analyzer for tool-calling agent execution traces."
+readme = "README.md"
+requires-python = ">=3.10"
+license = { file = "LICENSE" }
+authors = [{ name = "Ashwin Ugale" }]
+keywords = [
+    "agents", "llm", "tool-calling", "traces", "static-analysis",
+    "linter", "observability", "reliability",
+]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Intended Audience :: Developers",
+    "Topic :: Software Development :: Quality Assurance",
+    "Topic :: Software Development :: Testing",
+]
+
+# Light core: jsonschema + stdlib only. Heavy integrations live behind extras.
+dependencies = [
+    "jsonschema>=4.0",
+]
+
+[project.optional-dependencies]
+# A real OpenAI ReAct agent used to *generate* traces to lint (the system under test,
+# never part of the deterministic linter). Opt-in; the whole test/validation path is offline.
+real-agent = ["openai>=1.0"]
+# Everything needed to run the full test suite.
+dev = [
+    "pytest>=7",
+    "ruff>=0.4",
+]
+
+[project.scripts]
+tracelint = "tracelint.cli:main"
+
+[project.urls]
+Homepage = "https://github.com/AshwinUgale/tracelint"
+Repository = "https://github.com/AshwinUgale/tracelint"
+Issues = "https://github.com/AshwinUgale/tracelint/issues"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/tracelint"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-q"
+# Run the suite from a source checkout without an editable install.
+pythonpath = ["src"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py310"
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "UP", "B"]

tracelint-0.1.0/src/tracelint/__init__.py

@@ -0,0 +1,118 @@
+"""tracelint — a deterministic, judge-free static analyzer for tool-calling agent traces.
+
+The names re-exported here are the **supported public API** and follow semantic versioning.
+Everything else remains importable from its submodule (e.g. ``tracelint.rules``) but is internal
+and may change between minor versions.
+
+The linter reads a trace and emits structured *findings*; it never calls a model to judge. See
+``PROJECTS-TECHNICAL-SPEC.md`` Part II for the authoritative design.
+"""
+
+# ruff: noqa: I001 — imports grouped by role (matching __all__), not alphabetically.
+
+# --- Canonical trace schema ---------------------------------------------------------
+from tracelint.trace import (
+    Message,
+    ResultStatus,
+    Role,
+    Step,
+    StepMeta,
+    ToolCall,
+    ToolResult,
+    Trace,
+    build_trace,
+    load_traces,
+)
+
+# --- Tool ground truth --------------------------------------------------------------
+from tracelint.tools import ToolMetadata, ToolRegistry, ToolSpec
+
+# --- Findings + report --------------------------------------------------------------
+from tracelint.findings import ConfidenceTier, Finding, LintReport
+
+# --- Rules + driver -----------------------------------------------------------------
+from tracelint.rules import (
+    ErrorHandlingRule,
+    HallucinatedArgRule,
+    LoopRule,
+    RedundantCallRule,
+    Rule,
+    SchemaViolationRule,
+    ToolErrorEventRule,
+    default_rules,
+    lint_trace,
+)
+
+# --- Provenance ---------------------------------------------------------------------
+from tracelint.provenance import ProvenanceGraph, SourceType, build_provenance
+
+# --- Reliability: fault injection, statistics, nondeterminism -----------------------
+from tracelint.injection import (
+    FaultInjector,
+    FaultType,
+    RandomInjection,
+    TargetedInjection,
+    apply_fault,
+)
+from tracelint.stats import bootstrap_mean_ci, wilson_interval
+from tracelint.nondeterminism import (
+    FindingReproduction,
+    ReproductionReport,
+    aggregate_runs,
+    lint_runs,
+)
+
+# --- Recovery scorecard -------------------------------------------------------------
+from tracelint.scorecard import (
+    FaultRecovery,
+    Scorecard,
+    Task,
+    all_of,
+    final_answer_contains,
+    final_answer_not_claims,
+    render_scorecard,
+    run_scorecard,
+    state_check,
+    tool_called,
+)
+
+# --- Adapters -----------------------------------------------------------------------
+from tracelint.adapters import from_openai_messages, openai_tools_to_registry
+
+# --- Reporting ----------------------------------------------------------------------
+from tracelint.report import (
+    read_json,
+    render_html,
+    render_report,
+    render_reports,
+    write_html,
+    write_json,
+)
+
+__all__ = [
+    # Trace schema
+    "Trace", "Step", "Message", "ToolCall", "ToolResult", "StepMeta",
+    "Role", "ResultStatus", "build_trace", "load_traces",
+    # Tools
+    "ToolRegistry", "ToolSpec", "ToolMetadata",
+    # Findings
+    "Finding", "ConfidenceTier", "LintReport",
+    # Rules
+    "Rule", "lint_trace", "SchemaViolationRule", "ToolErrorEventRule", "ErrorHandlingRule",
+    "HallucinatedArgRule", "LoopRule", "RedundantCallRule", "default_rules",
+    # Provenance
+    "build_provenance", "ProvenanceGraph", "SourceType",
+    # Reliability
+    "FaultInjector", "FaultType", "TargetedInjection", "RandomInjection", "apply_fault",
+    "wilson_interval", "bootstrap_mean_ci",
+    "aggregate_runs", "lint_runs", "ReproductionReport", "FindingReproduction",
+    # Scorecard
+    "Task", "Scorecard", "FaultRecovery", "run_scorecard", "render_scorecard",
+    "tool_called", "final_answer_contains", "final_answer_not_claims", "state_check", "all_of",
+    # Adapters
+    "from_openai_messages", "openai_tools_to_registry",
+    # Reporting
+    "render_report", "render_reports", "render_html", "write_json", "write_html", "read_json",
+]
+
+__version__ = "0.1.0"