diff --git a/docs/testing/README.md b/docs/testing/README.md
new file mode 100644
index 00000000..be71ea1b
--- /dev/null
+++ b/docs/testing/README.md
@@ -0,0 +1,26 @@
+# Testing Docs
+
+Strongclaw keeps contributor-facing testing documentation in this directory.
+This is the canonical source of truth for how tests are organized, authored,
+and run inside the repository.
+
+Testing docs do not ship in `src/clawops/assets/...` unless a packaged runtime
+flow actually needs them.
+
+## Start Here
+
+- Read [authoring.md](authoring.md) when you are writing new tests, moving
+  existing coverage, or touching the pytest framework.
+- Read [operations.md](operations.md) when you are running tests locally,
+  debugging CI, or checking governance commands.
+- Read [../../tests/fixtures/README.md](../../tests/fixtures/README.md) for
+  fixture-package-specific guidance only.
+
+## Canonical Rules
+
+- Put tests under `tests/suites/{unit,integration,contracts,e2e,...}` and place
+  them in a subsystem-specific subdirectory when one exists.
+- Treat `TestContext` as the default path for patching, environment mutation,
+  and temporary working-directory changes.
+- Keep pytest framework rules and contributor docs together under
+  `docs/testing/` instead of scattering them across packaged asset trees.
diff --git a/docs/testing/authoring.md b/docs/testing/authoring.md
new file mode 100644
index 00000000..41cde00b
--- /dev/null
+++ b/docs/testing/authoring.md
@@ -0,0 +1,93 @@
+# Authoring Tests
+
+## Lane Model
+
+Strongclaw uses four primary default pytest lanes:
+
+- `unit`: isolated behavior and small-surface regression checks
+- `integration`: cross-module or service-shaped behavior
+- `contracts`: repository policies, docs parity, and CI/test-governance rules
+- `e2e`: black-box CLI and workflow-shaped orchestration coverage
+
+The repository also maintains an explicit `framework` lane for pytest-framework
+self-checks. Framework tests live under
+`tests/suites/contracts/testing/framework/` and are excluded from default runs.
+Use that lane only for pytest bootstrap and plugin-registration behavior.
+
+Monkeypatch governance is a default contract, not a framework-only self-check.
+The direct-monkeypatch contract lives under `tests/suites/contracts/testing/`
+so ordinary pytest runs fail when new unmanaged `monkeypatch` usage appears in
+suite code.
+
+Capability markers are additive and remain module-local:
+
+- `hypermemory`
+- `qdrant`
+- `network_local`
+
+Structural markers are assigned from the suite path layout in
+`tests/conftest.py`.
+
+## Placement Rules
+
+Add a new test in the suite that matches its behavior:
+
+- `tests/suites/unit/...` for isolated behavior
+- `tests/suites/integration/...` for cross-module or service-backed behavior
+- `tests/suites/contracts/...` for repository and governance rules
+- `tests/suites/contracts/testing/framework/...` for explicit pytest-framework
+  self-checks
+- `tests/suites/e2e/...` for black-box CLI and workflow-shaped coverage
+
+Prefer a dedicated subsystem directory inside each lane when one already
+exists.
+
+## Runtime Boundaries
+
+Use `tests/plugins/infrastructure/` for structural test runtime behavior such
+as `TestContext`, environment management, patch management, profile
+registration, and framework-owned pytest hooks.
+
+Use `tests/fixtures/` for domain-facing pytest fixture plugins only.
+
+Use `tests/utils/helpers/` for builders, subsystem runtimes, AST tooling, and
+other reusable support code that is not itself a fixture or pytest framework
+surface.
+
+Tests consume fixtures by name through pytest injection and should not import
+from `tests.fixtures`.
+
+## Preferred Authoring Path
+
+Environment mutation, working-directory changes, and patching should flow
+through the infrastructure runtime:
+
+- use `test_context.patch.patch(...)` or `patch_object(...)`
+- use `test_context.env.set(...)`, `remove(...)`, `update(...)`, or
+  `prepend_path(...)`
+- use `test_context.chdir(...)` for temporary working-directory changes
+
+Do not add new ordinary-suite tests that depend on raw `monkeypatch` unless the
+test is an explicit governed exception.
+
+## Service Resolution
+
+Service-backed tests resolve mode with this precedence:
+
+1. CLI: `--mock <service>`
+2. Environment: `<SERVICE>_TEST_MODE`
+3. Marker kwargs: `@pytest.mark.<service>(mode="real")`
+4. Default mode
+
+Current service support:
+
+- `qdrant`
+
+## Governance
+
+Framework policy lives under `tests/suites/contracts/testing/`.
+Add a contract test when a rule must stay true even if the implementation
+changes.
+
+Pytest-framework registration and bootstrap topology lives under
+`tests/suites/contracts/testing/framework/`.
diff --git a/docs/testing/operations.md b/docs/testing/operations.md
new file mode 100644
index 00000000..2ede7430
--- /dev/null
+++ b/docs/testing/operations.md
@@ -0,0 +1,42 @@
+# Testing Operations
+
+## Canonical Commands
+
+- Full suite: `uv run pytest -q`
+- Unit lane: `uv run pytest -q -m unit`
+- Integration lane: `uv run pytest -q -m integration`
+- Contract lane: `uv run pytest -q -m contract`
+- Framework lane only:
+  `uv run pytest -q -m framework tests/suites/contracts/testing/framework`
+- E2E lane: `uv run pytest -q -m e2e`
+- Hypermemory lane: `uv run pytest -q -m hypermemory`
+- Qdrant lane: `uv run pytest -q -m "hypermemory and qdrant"`
+
+## Service Mode Commands
+
+- Force Qdrant mock mode:
+  `uv run pytest -q -m "hypermemory and qdrant" --mock qdrant`
+- Force Qdrant real mode:
+  `QDRANT_TEST_MODE=real uv run pytest -q -m "hypermemory and qdrant"`
+- Provide an existing real endpoint:
+  `TEST_QDRANT_URL=http://127.0.0.1:6333 uv run pytest -q -m "hypermemory and qdrant"`
+- Override the managed Qdrant image:
+  `TEST_QDRANT_IMAGE=ghcr.io/example/qdrant:test uv run pytest -q -m "hypermemory and qdrant"`
+
+## Governance Checks
+
+- Testing contracts: `uv run pytest -q tests/suites/contracts/testing`
+- Fixture analysis:
+  `uv run python -m tests.utils.scripts.analyze_fixtures --json`
+- Safe timeout wrapper:
+  `uv run python -m tests.utils.scripts.pytest_safe --timeout 600 -q -m integration`
+
+## Common Triage
+
+- If the monkeypatch governance contract fails, migrate the test to
+  `TestContext` instead of expanding the allowlist by default.
+- If docs or links move, rerun the repository docs contracts so relative-link
+  and layout drift is caught early.
+- If a test needs reusable setup by fixture injection, add or extend a fixture.
+  If it needs reusable non-fixture logic, move that code into
+  `tests/utils/helpers/`.
diff --git a/platform/docs/TESTING_FRAMEWORK.md b/platform/docs/TESTING_FRAMEWORK.md
deleted file mode 100644
index 3b4a53d0..00000000
--- a/platform/docs/TESTING_FRAMEWORK.md
+++ /dev/null
@@ -1,111 +0,0 @@
-# Testing Framework
-
-## Lane Model
-
-Strongclaw uses four primary default pytest lanes:
-- `unit`: isolated behavior and small-surface regression checks
-- `integration`: cross-module or service-shaped behavior
-- `contracts`: repository policies, docs parity, and CI/test-governance rules
-- `e2e`: black-box CLI and workflow-shaped orchestration coverage
-
-The repository also maintains an explicit `framework` lane for pytest-framework self-checks.
-Framework tests live under `tests/suites/contracts/testing/framework/` and are excluded from
-default runs via the project pytest configuration. Run them explicitly when changing pytest
-bootstrap, plugin registration, or framework governance behavior.
-
-Capability markers are additive and remain module-local:
-- `hypermemory`
-- `qdrant`
-- `network_local`
-
-Structural markers are assigned from the suite path layout in `tests/conftest.py`.
-
-## Fixture and Helper Split
-
-Use `tests/plugins/infrastructure/` for structural test runtime behavior such as `TestContext`,
-environment management, patch management, profile registration, and framework-owned pytest hooks.
-Use `tests/fixtures/` for domain-facing pytest fixture plugins only.
-Use `tests/utils/helpers/` for builders, subsystem runtime helpers, AST tooling, and other
-non-structural support code.
-
-Keep root `tests/conftest.py` lean:
-- structural marker assignment
-- shared plugin registration
-- path fixtures that are meaningful across suites
-
-Root `tests/conftest.py` registers the infrastructure runtime and the shared fixture package via
-`pytest_plugins`.
-`tests/plugins/infrastructure/__init__.py` owns framework CLI options, universal `TestContext`,
-managed env injection, patch teardown, and named runtime profiles.
-`tests/fixtures/__init__.py` aggregates domain packages, and domain package `__init__.py` files
-aggregate their leaf fixture modules.
-Tests consume fixtures by name through pytest injection and should not import from `tests.fixtures`.
-Tests that need reusable builders, fakes, or types should import them from `tests.utils.helpers`.
-Environment mutation and patching should flow through the infrastructure runtime, for example
-`prepend_path`, `TestContext.env`, and `TestContext.patch`, instead of direct `monkeypatch`
-usage in ordinary suite code.
-
-## DualMode Service Resolution
-
-Service-backed tests resolve mode with this precedence:
-1. CLI: `--mock <service>`
-2. Environment: `<SERVICE>_TEST_MODE`
-3. Marker kwargs: `@pytest.mark.<service>(mode="real")`
-4. Default mode
-
-Current service support:
-- `qdrant`
-
-Examples:
-- `uv run pytest -q -m unit`
-- `uv run pytest -q -m "hypermemory and qdrant" --mock qdrant`
-- `uv run pytest -q -m e2e`
-- `uv run pytest -q -m framework tests/suites/contracts/testing/framework`
-- `QDRANT_TEST_MODE=real uv run pytest -q -m "hypermemory and qdrant"`
-
-## Adopted Patterns
-
-The repository intentionally adopts a small subset of the AIOA testing architecture:
-- worker-aware per-test identity
-- deterministic tracked cleanup through `TestContext`
-- universal `TestContext` creation for every test
-- infrastructure-owned environment and patch isolation
-- named infrastructure profiles for repeated runtime setup
-- narrow runtime helpers for network and Qdrant integration coverage
-- contract tests and static analysis for framework governance
-
-## Rejected Patterns
-
-The repository intentionally does not adopt these patterns at current scale:
-- plugin dependency graphs
-- a monolithic framework control plane in root `conftest.py`
-- registry-heavy data factories
-- pytest-embedded documentation generation
-- parallel-safety abstractions before xdist is an actual requirement
-
-## Growth Triggers
-
-Revisit the design if any of these become true:
-- more than one helper module grows past the line-count guardrail
-- three or more service runtimes need ordered lifecycle orchestration
-- xdist becomes part of the default CI matrix
-- fixture ownership becomes ambiguous across multiple domains
-
-## Governance Contracts
-
-Framework policy lives under `tests/suites/contracts/testing/`.
-Add a contract test when a rule must stay true even if the implementation changes.
-
-Pytest-framework registration and bootstrap topology lives under
-`tests/suites/contracts/testing/framework/`. Use that lane for assertions about recursive plugin
-registration, explicit framework-only behavior, and other tests that should not run in the default
-suite.
-
-Current governance covers:
-- root bootstrap shape
-- workflow pytest invocation policy
-- test-context cleanup invariants
-- environment and patch isolation
-- runtime helper behavior
-- fixture-analysis health checks
-- testing documentation presence
diff --git a/platform/docs/TESTING_OPERATIONS.md b/platform/docs/TESTING_OPERATIONS.md
deleted file mode 100644
index 47a5d1fd..00000000
--- a/platform/docs/TESTING_OPERATIONS.md
+++ /dev/null
@@ -1,46 +0,0 @@
-# Testing Operations
-
-## Canonical Commands
-
-- Full suite: `uv run pytest -q`
-- Unit lane: `uv run pytest -q -m unit`
-- Integration lane: `uv run pytest -q -m integration`
-- Contract lane: `uv run pytest -q -m contract`
-- Framework lane only: `uv run pytest -q -m framework tests/suites/contracts/testing/framework`
-- E2E lane: `uv run pytest -q -m e2e`
-- Hypermemory lane: `uv run pytest -q -m hypermemory`
-- Qdrant lane: `uv run pytest -q -m "hypermemory and qdrant"`
-
-## DualMode Commands
-
-- Force Qdrant mock mode: `uv run pytest -q -m "hypermemory and qdrant" --mock qdrant`
-- Force Qdrant real mode: `QDRANT_TEST_MODE=real uv run pytest -q -m "hypermemory and qdrant"`
-- Provide an existing real endpoint: `TEST_QDRANT_URL=http://127.0.0.1:6333 uv run pytest -q -m "hypermemory and qdrant"`
-- Override the managed Qdrant image: `TEST_QDRANT_IMAGE=ghcr.io/example/qdrant:test uv run pytest -q -m "hypermemory and qdrant"`
-  The default managed image is the repo-pinned official Qdrant GHCR mirror.
-
-## Governance Checks
-
-- Framework contracts: `uv run pytest -q tests/suites/contracts/testing`
-- Fixture analysis: `uv run python -m tests.utils.scripts.analyze_fixtures --json`
-- Safe timeout wrapper: `uv run python -m tests.utils.scripts.pytest_safe --timeout 600 -q -m integration`
-
-## Adding Coverage
-
-Add a new test in the suite that matches its behavior:
-- `tests/suites/unit/...` for isolated behavior
-- `tests/suites/integration/...` for cross-module or service-backed behavior
-- `tests/suites/contracts/...` for repository and governance rules
-- `tests/suites/contracts/testing/framework/...` for explicit pytest-framework self-checks
-- `tests/suites/e2e/...` for black-box CLI and workflow-shaped coverage
-
-Add a new fixture when pytest injection is the public entrypoint.
-Add a new helper when the logic should be reusable outside fixture setup.
-Import reusable support code from `tests.utils.helpers`, not from `tests.fixtures`.
-Use `prepend_path` or `test_context.env.prepend_path(...)` for PATH mutation.
-Use `test_context.patch.patch(...)` or `patch_object(...)` for managed patch teardown.
-
-Capability markers stay module-local.
-Structural markers come from the suite path layout and should not be added manually.
-Use `prepend_path`, `TestContext.env`, and `TestContext.patch` for environment and patch
-management instead of direct `monkeypatch` calls in ordinary suite code.
diff --git a/src/clawops/assets/platform/docs/TESTING_FRAMEWORK.md b/src/clawops/assets/platform/docs/TESTING_FRAMEWORK.md
deleted file mode 100644
index 3b4a53d0..00000000
--- a/src/clawops/assets/platform/docs/TESTING_FRAMEWORK.md
+++ /dev/null
@@ -1,111 +0,0 @@
-# Testing Framework
-
-## Lane Model
-
-Strongclaw uses four primary default pytest lanes:
-- `unit`: isolated behavior and small-surface regression checks
-- `integration`: cross-module or service-shaped behavior
-- `contracts`: repository policies, docs parity, and CI/test-governance rules
-- `e2e`: black-box CLI and workflow-shaped orchestration coverage
-
-The repository also maintains an explicit `framework` lane for pytest-framework self-checks.
-Framework tests live under `tests/suites/contracts/testing/framework/` and are excluded from
-default runs via the project pytest configuration. Run them explicitly when changing pytest
-bootstrap, plugin registration, or framework governance behavior.
-
-Capability markers are additive and remain module-local:
-- `hypermemory`
-- `qdrant`
-- `network_local`
-
-Structural markers are assigned from the suite path layout in `tests/conftest.py`.
-
-## Fixture and Helper Split
-
-Use `tests/plugins/infrastructure/` for structural test runtime behavior such as `TestContext`,
-environment management, patch management, profile registration, and framework-owned pytest hooks.
-Use `tests/fixtures/` for domain-facing pytest fixture plugins only.
-Use `tests/utils/helpers/` for builders, subsystem runtime helpers, AST tooling, and other
-non-structural support code.
-
-Keep root `tests/conftest.py` lean:
-- structural marker assignment
-- shared plugin registration
-- path fixtures that are meaningful across suites
-
-Root `tests/conftest.py` registers the infrastructure runtime and the shared fixture package via
-`pytest_plugins`.
-`tests/plugins/infrastructure/__init__.py` owns framework CLI options, universal `TestContext`,
-managed env injection, patch teardown, and named runtime profiles.
-`tests/fixtures/__init__.py` aggregates domain packages, and domain package `__init__.py` files
-aggregate their leaf fixture modules.
-Tests consume fixtures by name through pytest injection and should not import from `tests.fixtures`.
-Tests that need reusable builders, fakes, or types should import them from `tests.utils.helpers`.
-Environment mutation and patching should flow through the infrastructure runtime, for example
-`prepend_path`, `TestContext.env`, and `TestContext.patch`, instead of direct `monkeypatch`
-usage in ordinary suite code.
-
-## DualMode Service Resolution
-
-Service-backed tests resolve mode with this precedence:
-1. CLI: `--mock <service>`
-2. Environment: `<SERVICE>_TEST_MODE`
-3. Marker kwargs: `@pytest.mark.<service>(mode="real")`
-4. Default mode
-
-Current service support:
-- `qdrant`
-
-Examples:
-- `uv run pytest -q -m unit`
-- `uv run pytest -q -m "hypermemory and qdrant" --mock qdrant`
-- `uv run pytest -q -m e2e`
-- `uv run pytest -q -m framework tests/suites/contracts/testing/framework`
-- `QDRANT_TEST_MODE=real uv run pytest -q -m "hypermemory and qdrant"`
-
-## Adopted Patterns
-
-The repository intentionally adopts a small subset of the AIOA testing architecture:
-- worker-aware per-test identity
-- deterministic tracked cleanup through `TestContext`
-- universal `TestContext` creation for every test
-- infrastructure-owned environment and patch isolation
-- named infrastructure profiles for repeated runtime setup
-- narrow runtime helpers for network and Qdrant integration coverage
-- contract tests and static analysis for framework governance
-
-## Rejected Patterns
-
-The repository intentionally does not adopt these patterns at current scale:
-- plugin dependency graphs
-- a monolithic framework control plane in root `conftest.py`
-- registry-heavy data factories
-- pytest-embedded documentation generation
-- parallel-safety abstractions before xdist is an actual requirement
-
-## Growth Triggers
-
-Revisit the design if any of these become true:
-- more than one helper module grows past the line-count guardrail
-- three or more service runtimes need ordered lifecycle orchestration
-- xdist becomes part of the default CI matrix
-- fixture ownership becomes ambiguous across multiple domains
-
-## Governance Contracts
-
-Framework policy lives under `tests/suites/contracts/testing/`.
-Add a contract test when a rule must stay true even if the implementation changes.
-
-Pytest-framework registration and bootstrap topology lives under
-`tests/suites/contracts/testing/framework/`. Use that lane for assertions about recursive plugin
-registration, explicit framework-only behavior, and other tests that should not run in the default
-suite.
-
-Current governance covers:
-- root bootstrap shape
-- workflow pytest invocation policy
-- test-context cleanup invariants
-- environment and patch isolation
-- runtime helper behavior
-- fixture-analysis health checks
-- testing documentation presence
diff --git a/src/clawops/assets/platform/docs/TESTING_OPERATIONS.md b/src/clawops/assets/platform/docs/TESTING_OPERATIONS.md
deleted file mode 100644
index 47a5d1fd..00000000
--- a/src/clawops/assets/platform/docs/TESTING_OPERATIONS.md
+++ /dev/null
@@ -1,46 +0,0 @@
-# Testing Operations
-
-## Canonical Commands
-
-- Full suite: `uv run pytest -q`
-- Unit lane: `uv run pytest -q -m unit`
-- Integration lane: `uv run pytest -q -m integration`
-- Contract lane: `uv run pytest -q -m contract`
-- Framework lane only: `uv run pytest -q -m framework tests/suites/contracts/testing/framework`
-- E2E lane: `uv run pytest -q -m e2e`
-- Hypermemory lane: `uv run pytest -q -m hypermemory`
-- Qdrant lane: `uv run pytest -q -m "hypermemory and qdrant"`
-
-## DualMode Commands
-
-- Force Qdrant mock mode: `uv run pytest -q -m "hypermemory and qdrant" --mock qdrant`
-- Force Qdrant real mode: `QDRANT_TEST_MODE=real uv run pytest -q -m "hypermemory and qdrant"`
-- Provide an existing real endpoint: `TEST_QDRANT_URL=http://127.0.0.1:6333 uv run pytest -q -m "hypermemory and qdrant"`
-- Override the managed Qdrant image: `TEST_QDRANT_IMAGE=ghcr.io/example/qdrant:test uv run pytest -q -m "hypermemory and qdrant"`
-  The default managed image is the repo-pinned official Qdrant GHCR mirror.
-
-## Governance Checks
-
-- Framework contracts: `uv run pytest -q tests/suites/contracts/testing`
-- Fixture analysis: `uv run python -m tests.utils.scripts.analyze_fixtures --json`
-- Safe timeout wrapper: `uv run python -m tests.utils.scripts.pytest_safe --timeout 600 -q -m integration`
-
-## Adding Coverage
-
-Add a new test in the suite that matches its behavior:
-- `tests/suites/unit/...` for isolated behavior
-- `tests/suites/integration/...` for cross-module or service-backed behavior
-- `tests/suites/contracts/...` for repository and governance rules
-- `tests/suites/contracts/testing/framework/...` for explicit pytest-framework self-checks
-- `tests/suites/e2e/...` for black-box CLI and workflow-shaped coverage
-
-Add a new fixture when pytest injection is the public entrypoint.
-Add a new helper when the logic should be reusable outside fixture setup.
-Import reusable support code from `tests.utils.helpers`, not from `tests.fixtures`.
-Use `prepend_path` or `test_context.env.prepend_path(...)` for PATH mutation.
-Use `test_context.patch.patch(...)` or `patch_object(...)` for managed patch teardown.
-
-Capability markers stay module-local.
-Structural markers come from the suite path layout and should not be added manually.
-Use `prepend_path`, `TestContext.env`, and `TestContext.patch` for environment and patch
-management instead of direct `monkeypatch` calls in ordinary suite code.
diff --git a/tests/fixtures/README.md b/tests/fixtures/README.md
index bfe8d19c..7eced5e0 100644
--- a/tests/fixtures/README.md
+++ b/tests/fixtures/README.md
@@ -1,45 +1,27 @@
 # Test Fixture Layout
 
+Contributor-facing testing rules live under
+[`docs/testing/`](../../docs/testing/README.md).
+This file is only for fixture-package-specific guidance.
+
 `tests/plugins/infrastructure/` contains the structural runtime for every test.
-It owns the universal `TestContext`, framework env injection, patch teardown, profile handling,
-and infrastructure-owned pytest hooks.
+It owns the universal `TestContext`, framework env injection, tracked cwd
+changes, patch teardown, profile handling, and infrastructure-owned pytest
+hooks.
 
 `tests/fixtures/` contains domain-facing pytest activation surfaces.
-The package is loaded through root `tests/conftest.py`, then aggregated by package-level
-`pytest_plugins` registries:
+The package is loaded through root `tests/conftest.py`, then aggregated by
+package-level `pytest_plugins` registries:
+
 - `tests.plugins.infrastructure`
 - `tests.fixtures`
 - `tests.fixtures.core`
 - `tests.fixtures.platform`
 - `tests.fixtures.hypermemory`
 
-Leaf fixture modules should stay thin and expose fixtures that are meaningful to multiple tests or suites.
+Leaf fixture modules should stay thin and expose fixtures that are meaningful
+to multiple tests or suites.
 Tests should use fixture injection instead of importing from `tests.fixtures`.
 
-`tests/utils/helpers/` contains reusable builders, subsystem runtimes, and support code.
-Move implementation detail there when it is not itself structural infrastructure or a pytest fixture.
-Import helper functions, fakes, and types from `tests.utils.helpers.*`.
-
-Add a new fixture module when:
-- A set of related fixtures belongs to a clear subsystem boundary.
-- The fixture is reused across multiple tests or suites.
-- The module can stay focused without becoming a general grab bag.
-
-Add a new helper module when:
-- The code constructs data, manages lifecycle, or performs AST/runtime analysis.
-- The logic can be reused outside pytest fixture setup.
-- The module can be tested directly as a normal Python API.
-
-Add a contract test when:
-- A repository policy must not silently drift.
-- Fixture or helper placement matters for maintainability.
-- CI and local test invocation rules need enforcement.
-
-Three-layer pattern:
-- Infrastructure modules own universal test runtime behavior.
-- Fixture modules expose domain-facing pytest fixtures only.
-- Helper modules hold typed implementation logic and subsystem runtime helpers.
-
-Example:
-- `tests/fixtures/hypermemory/workspace.py` exports `hypermemory_workspace_factory`.
-- `tests/utils/helpers/hypermemory.py` implements the workspace builders and fake backends.
+Use [../../docs/testing/authoring.md](../../docs/testing/authoring.md) for
+lane placement, governance rules, and `TestContext` guidance.
diff --git a/tests/plugins/infrastructure/context.py b/tests/plugins/infrastructure/context.py
index e4d7d9f9..ab8a20ef 100644
--- a/tests/plugins/infrastructure/context.py
+++ b/tests/plugins/infrastructure/context.py
@@ -3,6 +3,7 @@
 from __future__ import annotations
 
 import logging
+import os
 import time
 from dataclasses import dataclass, field
 from typing import TYPE_CHECKING, Any, ClassVar, Protocol, cast, runtime_checkable
@@ -93,6 +94,7 @@ class TestContext:
     _cleaned: bool = False
     _environment: EnvironmentManager | None = field(default=None, init=False, repr=False)
     _patch_manager: PatchManager | None = field(default=None, init=False, repr=False)
+    _cwd_snapshot: str | None = field(default=None, init=False, repr=False)
 
     def attach_environment(self, manager: EnvironmentManager) -> None:
         """Bind the framework-managed environment runtime to this context."""
@@ -120,6 +122,13 @@ def apply_profiles(self, *names: str) -> None:
         """Apply one or more named runtime profiles to the current test."""
         self.env.apply_profiles(*names)
 
+    def chdir(self, path: str | os.PathLike[str]) -> None:
+        """Change the current working directory and restore it during cleanup."""
+        if self._cwd_snapshot is None:
+            self._cwd_snapshot = os.getcwd()
+            self.register_cleanup("cwd", self._restore_cwd)
+        os.chdir(path)
+
     def register_resource(
         self,
         name: str,
@@ -192,6 +201,12 @@ def _run_cleanup_stack(
                 errors.append((name, exc))
         return errors
 
+    def _restore_cwd(self) -> None:
+        if self._cwd_snapshot is None:
+            return
+        os.chdir(self._cwd_snapshot)
+        self._cwd_snapshot = None
+
 
 CONTEXT_KEY = pytest.StashKey[TestContext]()
 
diff --git a/tests/suites/contracts/repo/test_docs_parity.py b/tests/suites/contracts/repo/test_docs_parity.py
index 16c601e4..a7dfb8e0 100644
--- a/tests/suites/contracts/repo/test_docs_parity.py
+++ b/tests/suites/contracts/repo/test_docs_parity.py
@@ -18,12 +18,17 @@ def _official_markdown_files(repo_root: pathlib.Path) -> list[pathlib.Path]:
         repo_root / "SETUP_GUIDE.md",
         repo_root / "USAGE_GUIDE.md",
     ]
+    docs_markdown = (
+        [path for path in (repo_root / "docs").rglob("*.md") if path.is_file()]
+        if (repo_root / "docs").exists()
+        else []
+    )
     platform_markdown = [
         path
         for path in (repo_root / "platform").rglob("*.md")
         if "platform/plugins/" not in path.as_posix()
     ]
-    return root_markdown + platform_markdown
+    return root_markdown + docs_markdown + platform_markdown
 
 
 def test_markdown_relative_links_resolve() -> None:
diff --git a/tests/suites/contracts/repo/test_packaged_runtime_assets.py b/tests/suites/contracts/repo/test_packaged_runtime_assets.py
index c1251135..c49d8b78 100644
--- a/tests/suites/contracts/repo/test_packaged_runtime_assets.py
+++ b/tests/suites/contracts/repo/test_packaged_runtime_assets.py
@@ -9,7 +9,7 @@
 
 
 def _tracked_relative_files(root: pathlib.Path) -> set[pathlib.Path]:
-    """Return git-tracked files under one repo-relative root."""
+    """Return existing git-tracked files under one repo-relative root."""
     repo_relative_root = root.resolve().relative_to(REPO_ROOT)
     result = subprocess.run(
         [
@@ -29,7 +29,9 @@ def _tracked_relative_files(root: pathlib.Path) -> set[pathlib.Path]:
         if not raw_path:
             continue
         tracked_path = pathlib.Path(raw_path.decode("utf-8"))
-        files.add(tracked_path.relative_to(repo_relative_root))
+        relative_path = tracked_path.relative_to(repo_relative_root)
+        if (root / relative_path).is_file():
+            files.add(relative_path)
     return files
 
 
diff --git a/tests/suites/contracts/testing/framework/test_monkeypatch_allowlist.py b/tests/suites/contracts/testing/test_monkeypatch_governance.py
similarity index 93%
rename from tests/suites/contracts/testing/framework/test_monkeypatch_allowlist.py
rename to tests/suites/contracts/testing/test_monkeypatch_governance.py
index 270c3015..ab8af052 100644
--- a/tests/suites/contracts/testing/framework/test_monkeypatch_allowlist.py
+++ b/tests/suites/contracts/testing/test_monkeypatch_governance.py
@@ -1,4 +1,4 @@
-"""Framework governance for the remaining direct monkeypatch exceptions."""
+"""Governance for the remaining direct monkeypatch exceptions."""
 
 from __future__ import annotations
 
diff --git a/tests/suites/contracts/testing/test_test_context_contract.py b/tests/suites/contracts/testing/test_test_context_contract.py
index 880fe08d..7618c8d5 100644
--- a/tests/suites/contracts/testing/test_test_context_contract.py
+++ b/tests/suites/contracts/testing/test_test_context_contract.py
@@ -42,6 +42,7 @@ def test_test_context_fixture_is_accessed_from_stash(
 def test_test_context_always_exposes_env_and_patch_runtime(test_context: TestContext) -> None:
     assert test_context.env is not None
     assert test_context.patch is not None
+    assert callable(test_context.chdir)
 
 
 def test_fixture_registry_no_longer_loads_legacy_test_context_plugin() -> None:
diff --git a/tests/suites/contracts/testing/test_testing_docs_contract.py b/tests/suites/contracts/testing/test_testing_docs_contract.py
index 59d87ed4..4596f2f9 100644
--- a/tests/suites/contracts/testing/test_testing_docs_contract.py
+++ b/tests/suites/contracts/testing/test_testing_docs_contract.py
@@ -4,33 +4,44 @@
 
 from tests.utils.helpers.repo import REPO_ROOT
 
+DOCS_ROOT = REPO_ROOT / "docs" / "testing"
 
-def test_testing_framework_doc_exists() -> None:
-    assert (REPO_ROOT / "platform" / "docs" / "TESTING_FRAMEWORK.md").is_file()
+
+def test_testing_docs_entrypoint_exists() -> None:
+    assert (DOCS_ROOT / "README.md").is_file()
+
+
+def test_testing_authoring_doc_exists() -> None:
+    assert (DOCS_ROOT / "authoring.md").is_file()
 
 
 def test_testing_operations_doc_exists() -> None:
-    assert (REPO_ROOT / "platform" / "docs" / "TESTING_OPERATIONS.md").is_file()
+    assert (DOCS_ROOT / "operations.md").is_file()
 
 
 def test_fixture_readme_exists() -> None:
     assert (REPO_ROOT / "tests" / "fixtures" / "README.md").is_file()
 
 
-def test_docs_mention_lane_model() -> None:
-    framework_doc = (REPO_ROOT / "platform" / "docs" / "TESTING_FRAMEWORK.md").read_text(
-        encoding="utf-8"
-    )
+def test_docs_entrypoint_links_to_authoring_and_operations() -> None:
+    readme = (DOCS_ROOT / "README.md").read_text(encoding="utf-8")
+
+    assert "authoring.md" in readme
+    assert "operations.md" in readme
+    assert "tests/fixtures/README.md" in readme
+
+
+def test_authoring_doc_mentions_lane_model() -> None:
+    authoring_doc = (DOCS_ROOT / "authoring.md").read_text(encoding="utf-8")
 
     for lane in ("unit", "integration", "contracts", "framework"):
-        assert lane in framework_doc
+        assert lane in authoring_doc
 
 
 def test_docs_mention_infrastructure_runtime_namespace() -> None:
-    framework_doc = (REPO_ROOT / "platform" / "docs" / "TESTING_FRAMEWORK.md").read_text(
-        encoding="utf-8"
-    )
+    authoring_doc = (DOCS_ROOT / "authoring.md").read_text(encoding="utf-8")
     fixture_readme = (REPO_ROOT / "tests" / "fixtures" / "README.md").read_text(encoding="utf-8")
 
-    assert "tests/plugins/infrastructure" in framework_doc
+    assert "tests/plugins/infrastructure" in authoring_doc
+    assert "docs/testing/authoring.md" in fixture_readme
     assert "tests/plugins/infrastructure" in fixture_readme
diff --git a/tests/suites/integration/clawops/test_devflow_recovery.py b/tests/suites/integration/clawops/test_devflow_recovery.py
index 243d7961..cbff454b 100644
--- a/tests/suites/integration/clawops/test_devflow_recovery.py
+++ b/tests/suites/integration/clawops/test_devflow_recovery.py
@@ -11,6 +11,7 @@
 from clawops.devflow import main
 from clawops.devflow_roles import WorkspaceMode
 from clawops.devflow_workspaces import DevflowWorkspacePlanner, PlannedWorkspace
+from tests.plugins.infrastructure.context import TestContext
 from tests.utils.helpers.cli import PathPrepender
 from tests.utils.helpers.devflow import (
     init_git_repo,
@@ -70,7 +71,7 @@ def test_devflow_workspace_failure_marks_run_failed_and_audit_still_works(
     tmp_path: pathlib.Path,
     prepend_path: PathPrepender,
     capsys: pytest.CaptureFixture[str],
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     repo_root = tmp_path / "repo"
     repo_root.mkdir()
@@ -100,7 +101,7 @@ def _flaky_prepare(
             source_root=source_root,
         )
 
-    monkeypatch.setattr(DevflowWorkspacePlanner, "prepare", _flaky_prepare)
+    test_context.patch.patch_object(DevflowWorkspacePlanner, "prepare", new=_flaky_prepare)
 
     exit_code = main(["run", "--project-root", str(repo_root), "--goal", "workspace recovery"])
     payload = json.loads(capsys.readouterr().out)
diff --git a/tests/suites/unit/ci/test_memory_plugin_verification.py b/tests/suites/unit/ci/test_memory_plugin_verification.py
index 696c0244..322eae3a 100644
--- a/tests/suites/unit/ci/test_memory_plugin_verification.py
+++ b/tests/suites/unit/ci/test_memory_plugin_verification.py
@@ -5,8 +5,6 @@
 from pathlib import Path
 from typing import Any
 
-import pytest
-
 from tests.plugins.infrastructure.context import TestContext
 from tests.scripts import memory_plugin_verification as memory_plugin_script
 from tests.utils.helpers import ci_workflows
@@ -14,7 +12,7 @@
 
 
 def test_run_vendored_host_checks_installs_cli_and_clears_aws_env(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
     tmp_path: Path,
 ) -> None:
     """Vendored host checks should prefix PATH and clear ambient AWS variables."""
@@ -35,10 +33,10 @@ def fake_run_checked(
         seen_calls.append((command, cwd, env))
         return None
 
-    monkeypatch.setattr(memory_plugin_helpers, "run_checked", fake_run_checked)
-    monkeypatch.setenv("PATH", "/usr/bin")
-    monkeypatch.setenv("AWS_PROFILE", "default")
-    monkeypatch.setenv("AWS_REGION", "us-east-1")
+    test_context.patch.patch_object(memory_plugin_helpers, "run_checked", new=fake_run_checked)
+    test_context.env.set("PATH", "/usr/bin")
+    test_context.env.set("AWS_PROFILE", "default")
+    test_context.env.set("AWS_REGION", "us-east-1")
 
     ci_workflows.run_vendored_host_checks(repo_root)
 
@@ -55,7 +53,7 @@ def fake_run_checked(
     assert npm_ci_env["PATH"].startswith(str(Path(install_command[3]) / "node_modules" / ".bin"))
 
 
-def test_wait_for_qdrant_retries_until_ready(monkeypatch: pytest.MonkeyPatch) -> None:
+def test_wait_for_qdrant_retries_until_ready(test_context: TestContext) -> None:
     """Qdrant readiness should retry transient probe failures."""
     attempts: list[str] = []
     sleeps: list[float] = []
@@ -77,8 +75,12 @@ def fake_urlopen(url: str, timeout: int) -> _Response:
             raise OSError("not ready")
         return _Response()
 
-    monkeypatch.setattr(memory_plugin_helpers.urllib.request, "urlopen", fake_urlopen)
-    monkeypatch.setattr(memory_plugin_helpers.time, "sleep", sleeps.append)
+    test_context.patch.patch_object(
+        memory_plugin_helpers.urllib.request,
+        "urlopen",
+        new=fake_urlopen,
+    )
+    test_context.patch.patch_object(memory_plugin_helpers.time, "sleep", new=sleeps.append)
 
     ci_workflows.wait_for_qdrant("http://127.0.0.1:6333/healthz", attempts=4, sleep_seconds=1.5)
 
diff --git a/tests/suites/unit/ci/test_release_workflow.py b/tests/suites/unit/ci/test_release_workflow.py
index 23be4d68..21e27cbf 100644
--- a/tests/suites/unit/ci/test_release_workflow.py
+++ b/tests/suites/unit/ci/test_release_workflow.py
@@ -5,8 +5,6 @@
 from pathlib import Path
 from typing import Any
 
-import pytest
-
 from tests.plugins.infrastructure.context import TestContext
 from tests.utils.helpers import ci_workflows
 from tests.utils.helpers._ci_workflows import release as release_helpers
@@ -28,7 +26,7 @@ def test_clean_artifact_directories_removes_paths(tmp_path: Path) -> None:
 
 
 def test_verify_release_artifacts_runs_twine_and_smoke_tests(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
     tmp_path: Path,
 ) -> None:
     """Release verification should check artifacts and install both wheel and sdist."""
@@ -52,7 +50,7 @@ def fake_run_checked(
         seen_commands.append(command)
         return None
 
-    monkeypatch.setattr(release_helpers, "run_checked", fake_run_checked)
+    test_context.patch.patch_object(release_helpers, "run_checked", new=fake_run_checked)
 
     ci_workflows.verify_release_artifacts(dist_dir)
 
@@ -66,7 +64,7 @@ def fake_run_checked(
 
 
 def test_publish_github_release_creates_when_missing(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
     tmp_path: Path,
 ) -> None:
     """Release publishing should create the release when it does not exist yet."""
@@ -91,7 +89,7 @@ def fake_run_checked(
             raise ci_workflows.CiWorkflowError("missing release")
         return None
 
-    monkeypatch.setattr(release_helpers, "run_checked", fake_run_checked)
+    test_context.patch.patch_object(release_helpers, "run_checked", new=fake_run_checked)
 
     ci_workflows.publish_github_release("v1.0.0", dist_dir, sbom_path)
 
diff --git a/tests/suites/unit/clawops/assets/test_runtime_assets.py b/tests/suites/unit/clawops/assets/test_runtime_assets.py
index acb91b48..84dd1682 100644
--- a/tests/suites/unit/clawops/assets/test_runtime_assets.py
+++ b/tests/suites/unit/clawops/assets/test_runtime_assets.py
@@ -4,17 +4,16 @@
 
 import pathlib
 
-import pytest
-
 from clawops.runtime_assets import PACKAGED_ASSET_ROOT, resolve_asset_path, resolve_runtime_layout
+from tests.plugins.infrastructure.context import TestContext
 from tests.utils.helpers.repo import REPO_ROOT
 
 
 def test_runtime_layout_uses_packaged_assets_outside_source_checkout(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
-    monkeypatch.chdir(tmp_path)
+    test_context.chdir(tmp_path)
 
     layout = resolve_runtime_layout(home_dir=tmp_path / "home")
 
diff --git a/tests/suites/unit/clawops/cli/test_root_defaults.py b/tests/suites/unit/clawops/cli/test_root_defaults.py
index 4adae3fc..49c2ae2e 100644
--- a/tests/suites/unit/clawops/cli/test_root_defaults.py
+++ b/tests/suites/unit/clawops/cli/test_root_defaults.py
@@ -13,6 +13,7 @@
 import clawops.openclaw_config as openclaw_config
 import clawops.strongclaw_baseline as strongclaw_baseline
 from clawops.root_detection import resolve_project_root, resolve_strongclaw_repo_root
+from tests.plugins.infrastructure.context import TestContext
 
 
 def _init_strongclaw_repo(repo_root: pathlib.Path) -> pathlib.Path:
@@ -51,8 +52,8 @@ def test_resolve_project_root_prefers_the_nearest_git_ancestor(tmp_path: pathlib
 
 def test_render_openclaw_config_main_infers_repo_root_from_cwd(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
     capsys: pytest.CaptureFixture[str],
+    test_context: TestContext,
 ) -> None:
     repo_root = _init_strongclaw_repo(tmp_path / "repo")
     nested = repo_root / "platform" / "docs"
@@ -82,12 +83,16 @@ def _materialize_runtime_memory_configs(
         del home_dir, user_timezone
         return repo_root / "managed-memory.yaml", repo_root / "managed-memory.sqlite.yaml"
 
-    monkeypatch.chdir(nested)
-    monkeypatch.setattr(openclaw_config, "render_openclaw_profile", _render_openclaw_profile)
-    monkeypatch.setattr(
+    test_context.chdir(nested)
+    test_context.patch.patch_object(
+        openclaw_config,
+        "render_openclaw_profile",
+        new=_render_openclaw_profile,
+    )
+    test_context.patch.patch_object(
         openclaw_config,
         "materialize_runtime_memory_configs",
-        _materialize_runtime_memory_configs,
+        new=_materialize_runtime_memory_configs,
     )
 
     exit_code = openclaw_config.main(["--profile", "hypermemory", "--output", str(output_path)])
@@ -100,8 +105,8 @@ def _materialize_runtime_memory_configs(
 
 def test_strongclaw_baseline_infers_repo_root_and_runs_dir_from_cwd(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
     capsys: pytest.CaptureFixture[str],
+    test_context: TestContext,
 ) -> None:
     repo_root = _init_strongclaw_repo(tmp_path / "repo")
     nested = repo_root / "src" / "clawops"
@@ -117,8 +122,8 @@ def _verify_baseline(
         recorded_runs_dir = runs_dir
         return {"ok": True, "runsDir": str(runs_dir)}
 
-    monkeypatch.chdir(nested)
-    monkeypatch.setattr(strongclaw_baseline, "verify_baseline", _verify_baseline)
+    test_context.chdir(nested)
+    test_context.patch.patch_object(strongclaw_baseline, "verify_baseline", new=_verify_baseline)
 
     exit_code = strongclaw_baseline.main(["verify"])
 
@@ -130,7 +135,7 @@ def _verify_baseline(
 
 def test_acp_runner_infers_project_root_without_explicit_flag(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     project_root = tmp_path / "project"
     (project_root / ".git").mkdir(parents=True)
@@ -140,7 +145,7 @@ def test_acp_runner_infers_project_root_without_explicit_flag(
     workspace.mkdir()
     state_dir = tmp_path / "state"
 
-    monkeypatch.chdir(nested)
+    test_context.chdir(nested)
 
     args = acp_runner.parse_args(
         [
diff --git a/tests/suites/unit/clawops/cli/test_root_flag_aliases.py b/tests/suites/unit/clawops/cli/test_root_flag_aliases.py
index 75b948d8..d5c8bf84 100644
--- a/tests/suites/unit/clawops/cli/test_root_flag_aliases.py
+++ b/tests/suites/unit/clawops/cli/test_root_flag_aliases.py
@@ -17,6 +17,7 @@
     resolve_source_root_argument,
     warn_ignored_repo_root_argument,
 )
+from tests.plugins.infrastructure.context import TestContext
 
 
 def _init_source_checkout(root: pathlib.Path) -> pathlib.Path:
@@ -64,11 +65,11 @@ def test_project_root_legacy_repo_root_alias_warns(
 
 def test_source_root_requires_source_root_guidance_when_not_discoverable(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     parser = argparse.ArgumentParser()
     add_source_root_argument(parser)
-    monkeypatch.chdir(tmp_path)
+    test_context.chdir(tmp_path)
 
     with pytest.raises(FileNotFoundError, match="pass --source-root explicitly\\."):
         resolve_source_root_argument(parser.parse_args([]), command_name="clawops baseline")
diff --git a/tests/suites/integration/clawops/test_setup_runtime_boundaries.py b/tests/suites/unit/clawops/cli/test_setup_runtime_boundaries.py
similarity index 71%
rename from tests/suites/integration/clawops/test_setup_runtime_boundaries.py
rename to tests/suites/unit/clawops/cli/test_setup_runtime_boundaries.py
index ff2fd96b..96b12a6f 100644
--- a/tests/suites/integration/clawops/test_setup_runtime_boundaries.py
+++ b/tests/suites/unit/clawops/cli/test_setup_runtime_boundaries.py
@@ -1,4 +1,4 @@
-"""Integration coverage for setup/doctor runtime-boundary behavior."""
+"""Unit coverage for setup/doctor runtime-boundary behavior."""
 
 from __future__ import annotations
 
@@ -8,11 +8,12 @@
 
 from clawops import cli as root_cli
 from clawops import setup_cli
+from tests.plugins.infrastructure.context import TestContext
 
 
 def test_root_cli_setup_render_only_path_skips_model_auth(
-    monkeypatch: pytest.MonkeyPatch,
     tmp_path: pathlib.Path,
+    test_context: TestContext,
 ) -> None:
     """The public CLI should not require model auth on a render-only setup path."""
     calls: list[str] = []
@@ -74,13 +75,17 @@ def _render_service_files(repo_root: pathlib.Path) -> dict[str, object]:
         calls.append("services-render")
         return {"ok": True}
 
-    monkeypatch.setattr(setup_cli, "bootstrap_state_ready", _bootstrap_state_ready)
-    monkeypatch.setattr(setup_cli, "install_profile_assets", _install_profile_assets)
-    monkeypatch.setattr(setup_cli, "configure_varlock_env", _configure_varlock_env)
-    monkeypatch.setattr(setup_cli, "_render_openclaw_config", _render_openclaw_config)
-    monkeypatch.setattr(setup_cli, "_doctor_host_payload", _doctor_host_payload)
-    monkeypatch.setattr(setup_cli, "ensure_model_auth", _ensure_model_auth)
-    monkeypatch.setattr(setup_cli, "render_service_files", _render_service_files)
+    test_context.patch.patch_object(setup_cli, "bootstrap_state_ready", new=_bootstrap_state_ready)
+    test_context.patch.patch_object(
+        setup_cli, "install_profile_assets", new=_install_profile_assets
+    )
+    test_context.patch.patch_object(setup_cli, "configure_varlock_env", new=_configure_varlock_env)
+    test_context.patch.patch_object(
+        setup_cli, "_render_openclaw_config", new=_render_openclaw_config
+    )
+    test_context.patch.patch_object(setup_cli, "_doctor_host_payload", new=_doctor_host_payload)
+    test_context.patch.patch_object(setup_cli, "ensure_model_auth", new=_ensure_model_auth)
+    test_context.patch.patch_object(setup_cli, "render_service_files", new=_render_service_files)
 
     exit_code = root_cli.main(["setup", "--asset-root", str(tmp_path), "--no-activate-services"])
 
@@ -95,9 +100,9 @@ def _render_service_files(repo_root: pathlib.Path) -> dict[str, object]:
 
 
 def test_root_cli_doctor_bounded_path_skips_openclaw_runtime_audits(
-    monkeypatch: pytest.MonkeyPatch,
     tmp_path: pathlib.Path,
     capsys: pytest.CaptureFixture[str],
+    test_context: TestContext,
 ) -> None:
     """The public CLI should keep bounded doctor local when both skip flags are set."""
 
@@ -153,13 +158,15 @@ def _verify_channels(**kwargs: object) -> _OkReport:
         del kwargs
         return _OkReport()
 
-    monkeypatch.setattr(setup_cli, "configure_varlock_env", _configure_varlock_env)
-    monkeypatch.setattr(setup_cli, "_doctor_host_payload", _doctor_host_payload)
-    monkeypatch.setattr(setup_cli, "_require_model_check_ok", _require_model_check_ok)
-    monkeypatch.setattr(setup_cli, "run_openclaw_command", _run_openclaw_command)
-    monkeypatch.setattr(setup_cli, "verify_sidecars", _verify_sidecars)
-    monkeypatch.setattr(setup_cli, "verify_observability", _verify_observability)
-    monkeypatch.setattr(setup_cli, "verify_channels", _verify_channels)
+    test_context.patch.patch_object(setup_cli, "configure_varlock_env", new=_configure_varlock_env)
+    test_context.patch.patch_object(setup_cli, "_doctor_host_payload", new=_doctor_host_payload)
+    test_context.patch.patch_object(
+        setup_cli, "_require_model_check_ok", new=_require_model_check_ok
+    )
+    test_context.patch.patch_object(setup_cli, "run_openclaw_command", new=_run_openclaw_command)
+    test_context.patch.patch_object(setup_cli, "verify_sidecars", new=_verify_sidecars)
+    test_context.patch.patch_object(setup_cli, "verify_observability", new=_verify_observability)
+    test_context.patch.patch_object(setup_cli, "verify_channels", new=_verify_channels)
 
     exit_code = root_cli.main(
         ["doctor", "--asset-root", str(tmp_path), "--skip-runtime", "--no-model-probe"]
diff --git a/tests/suites/unit/clawops/context/test_codebase_graph.py b/tests/suites/unit/clawops/context/test_codebase_graph.py
index 27d82180..e03e65fd 100644
--- a/tests/suites/unit/clawops/context/test_codebase_graph.py
+++ b/tests/suites/unit/clawops/context/test_codebase_graph.py
@@ -13,6 +13,7 @@
     normalize_neo4j_driver_url,
     service_from_config,
 )
+from tests.plugins.infrastructure.context import TestContext
 
 
 def test_large_scale_requires_healthy_neo4j_even_when_fallback_is_allowed(
@@ -146,7 +147,7 @@ def test_medium_scale_symbol_graph_expands_related_files(
 
 
 def test_neo4j_neighbors_use_literal_validated_depth(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     backend = Neo4jGraphBackend(GraphConfig())
     recorded_query = ""
@@ -162,7 +163,7 @@ def _fake_run_query(
         recorded_parameters = parameters
         return [{"id": "neighbor-1"}]
 
-    monkeypatch.setattr(backend, "_run_query", _fake_run_query)
+    test_context.patch.patch_object(backend, "_run_query", new=_fake_run_query)
 
     neighbors = backend.neighbors(
         node_id="symbol:provider.rotate_token",
diff --git a/tests/suites/unit/clawops/context/test_codebase_hybrid.py b/tests/suites/unit/clawops/context/test_codebase_hybrid.py
index 24fc1620..45168991 100644
--- a/tests/suites/unit/clawops/context/test_codebase_hybrid.py
+++ b/tests/suites/unit/clawops/context/test_codebase_hybrid.py
@@ -6,13 +6,13 @@
 import uuid
 from collections.abc import Sequence
 
-import pytest
 import requests
 
 from clawops.common import write_yaml
 from clawops.context.codebase.service import CodebaseContextService, service_from_config
 from clawops.hypermemory.contracts import SparseVectorPayload, VectorPoint
 from clawops.hypermemory.models import DenseSearchCandidate, RerankResponse, SparseSearchCandidate
+from tests.plugins.infrastructure.context import TestContext
 
 
 class _FakeEmbeddingProvider:
@@ -258,7 +258,7 @@ def test_medium_scale_worker_syncs_chunk_vectors_when_hybrid_enabled(
 
 def test_medium_scale_worker_skips_sparse_rebuild_for_unchanged_warm_consolidation(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     repo, service = _build_service(tmp_path)
     (repo / "auth.py").write_text(
@@ -283,9 +283,9 @@ def _unexpected_sparse_rebuild(*args: object, **kwargs: object) -> object:
         del args, kwargs
         raise AssertionError("warm consolidation should reuse the persisted sparse state")
 
-    monkeypatch.setattr(
+    test_context.patch.patch(
         "clawops.context.codebase.service.build_sparse_encoder_from_documents",
-        _unexpected_sparse_rebuild,
+        new=_unexpected_sparse_rebuild,
     )
 
     service.consolidate_runtime_artifacts()
diff --git a/tests/suites/unit/clawops/test_strongclaw_baseline.py b/tests/suites/unit/clawops/test_strongclaw_baseline.py
index f592d7ea..31a6bfab 100644
--- a/tests/suites/unit/clawops/test_strongclaw_baseline.py
+++ b/tests/suites/unit/clawops/test_strongclaw_baseline.py
@@ -9,6 +9,7 @@
 
 from clawops import strongclaw_baseline
 from clawops.strongclaw_runtime import CommandError
+from tests.plugins.infrastructure.context import TestContext
 
 
 class _FakeCommandResult:
@@ -50,8 +51,8 @@ def _noop_harness_smoke(_repo: pathlib.Path, _runs_dir: pathlib.Path) -> None:
 
 
 def test_verify_baseline_uses_uv_dependency_group_for_repo_tests(
-    monkeypatch: pytest.MonkeyPatch,
     tmp_path: pathlib.Path,
+    test_context: TestContext,
 ) -> None:
     repo_root = _init_source_checkout(tmp_path / "repo")
     config_path = tmp_path / "openclaw.json"
@@ -99,29 +100,31 @@ def _run_harness_smoke(repo: pathlib.Path, runs_dir: pathlib.Path) -> None:
         assert repo == repo_root
         assert runs_dir == tmp_path / "runs"
 
-    monkeypatch.setattr(strongclaw_baseline, "require_openclaw", _require_openclaw)
-    monkeypatch.setattr(
+    test_context.patch.patch_object(strongclaw_baseline, "require_openclaw", new=_require_openclaw)
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "resolve_openclaw_config_path",
-        _resolve_openclaw_config_path,
+        new=_resolve_openclaw_config_path,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "run_openclaw_command",
-        _run_openclaw_command,
+        new=_run_openclaw_command,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "ensure_model_auth",
-        _ensure_model_auth,
+        new=_ensure_model_auth,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "rendered_openclaw_uses_hypermemory",
-        _rendered_openclaw_uses_hypermemory,
+        new=_rendered_openclaw_uses_hypermemory,
+    )
+    test_context.patch.patch_object(strongclaw_baseline, "run_command", new=_run_command)
+    test_context.patch.patch_object(
+        strongclaw_baseline, "run_harness_smoke", new=_run_harness_smoke
     )
-    monkeypatch.setattr(strongclaw_baseline, "run_command", _run_command)
-    monkeypatch.setattr(strongclaw_baseline, "run_harness_smoke", _run_harness_smoke)
 
     payload = strongclaw_baseline.verify_baseline(repo_root, runs_dir=tmp_path / "runs")
 
@@ -134,8 +137,8 @@ def _run_harness_smoke(repo: pathlib.Path, runs_dir: pathlib.Path) -> None:
 
 
 def test_verify_baseline_surfaces_repo_test_failure_detail(
-    monkeypatch: pytest.MonkeyPatch,
     tmp_path: pathlib.Path,
+    test_context: TestContext,
 ) -> None:
     repo_root = _init_source_checkout(tmp_path / "repo")
     config_path = tmp_path / "openclaw.json"
@@ -178,29 +181,31 @@ def _run_command(
             return _FakeCommandResult(ok=False, stderr="repo tests failed")
         return _FakeCommandResult(ok=True)
 
-    monkeypatch.setattr(strongclaw_baseline, "require_openclaw", _require_openclaw)
-    monkeypatch.setattr(
+    test_context.patch.patch_object(strongclaw_baseline, "require_openclaw", new=_require_openclaw)
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "resolve_openclaw_config_path",
-        _resolve_openclaw_config_path,
+        new=_resolve_openclaw_config_path,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "run_openclaw_command",
-        _run_openclaw_command,
+        new=_run_openclaw_command,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "ensure_model_auth",
-        _ensure_model_auth,
+        new=_ensure_model_auth,
     )
-    monkeypatch.setattr(
+    test_context.patch.patch_object(
         strongclaw_baseline,
         "rendered_openclaw_uses_hypermemory",
-        _rendered_openclaw_uses_hypermemory,
+        new=_rendered_openclaw_uses_hypermemory,
+    )
+    test_context.patch.patch_object(strongclaw_baseline, "run_command", new=_run_command)
+    test_context.patch.patch_object(
+        strongclaw_baseline, "run_harness_smoke", new=_noop_harness_smoke
     )
-    monkeypatch.setattr(strongclaw_baseline, "run_command", _run_command)
-    monkeypatch.setattr(strongclaw_baseline, "run_harness_smoke", _noop_harness_smoke)
 
     with pytest.raises(CommandError, match="repo tests failed"):
         strongclaw_baseline.verify_baseline(repo_root, runs_dir=tmp_path / "runs")
diff --git a/tests/suites/unit/clawops/test_strongclaw_runtime.py b/tests/suites/unit/clawops/test_strongclaw_runtime.py
index 30b0449c..99913756 100644
--- a/tests/suites/unit/clawops/test_strongclaw_runtime.py
+++ b/tests/suites/unit/clawops/test_strongclaw_runtime.py
@@ -8,6 +8,7 @@
 import clawops.strongclaw_runtime as runtime
 from clawops.app_paths import strongclaw_varlock_dir
 from clawops.strongclaw_runtime import CommandError, ExecResult, write_env_assignments
+from tests.plugins.infrastructure.context import TestContext
 
 
 def test_write_env_assignments_uses_owner_only_permissions(tmp_path: pathlib.Path) -> None:
@@ -21,7 +22,7 @@ def test_write_env_assignments_uses_owner_only_permissions(tmp_path: pathlib.Pat
 
 
 def test_docker_backend_diagnostics_capture_runtime_context(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
         if argv == ["docker", "context", "show"]:
@@ -36,11 +37,15 @@ def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
             )
         raise AssertionError(argv)
 
-    monkeypatch.setattr(runtime, "docker_cli_installed", lambda: True)
-    monkeypatch.setattr(runtime, "docker_compose_available", lambda: True)
-    monkeypatch.setattr(runtime, "detect_docker_runtime_provider", lambda: "OrbStack")
-    monkeypatch.setattr(runtime, "run_command", _fake_run_command)
-    monkeypatch.setenv("DOCKER_HOST", "unix:///Users/test/.orbstack/run/docker.sock")
+    test_context.patch.patch_object(runtime, "docker_cli_installed", new=lambda: True)
+    test_context.patch.patch_object(runtime, "docker_compose_available", new=lambda: True)
+    test_context.patch.patch_object(
+        runtime,
+        "detect_docker_runtime_provider",
+        new=lambda: "OrbStack",
+    )
+    test_context.patch.patch_object(runtime, "run_command", new=_fake_run_command)
+    test_context.env.set("DOCKER_HOST", "unix:///Users/test/.orbstack/run/docker.sock")
 
     diagnostics = runtime.docker_backend_diagnostics()
 
@@ -52,7 +57,7 @@ def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
 
 
 def test_ensure_docker_backend_ready_surfaces_runtime_details(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
         if argv == ["docker", "context", "show"]:
@@ -67,12 +72,16 @@ def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
             )
         raise AssertionError(argv)
 
-    monkeypatch.setattr(runtime, "docker_cli_installed", lambda: True)
-    monkeypatch.setattr(runtime, "docker_compose_available", lambda: True)
-    monkeypatch.setattr(runtime, "detect_docker_runtime_provider", lambda: "OrbStack")
-    monkeypatch.setattr(runtime, "run_command", _fake_run_command)
-    monkeypatch.setattr(runtime, "load_docker_refresh_state", lambda: None)
-    monkeypatch.setenv("DOCKER_HOST", "unix:///Users/test/.orbstack/run/docker.sock")
+    test_context.patch.patch_object(runtime, "docker_cli_installed", new=lambda: True)
+    test_context.patch.patch_object(runtime, "docker_compose_available", new=lambda: True)
+    test_context.patch.patch_object(
+        runtime,
+        "detect_docker_runtime_provider",
+        new=lambda: "OrbStack",
+    )
+    test_context.patch.patch_object(runtime, "run_command", new=_fake_run_command)
+    test_context.patch.patch_object(runtime, "load_docker_refresh_state", new=lambda: None)
+    test_context.env.set("DOCKER_HOST", "unix:///Users/test/.orbstack/run/docker.sock")
 
     with pytest.raises(CommandError) as exc_info:
         runtime.ensure_docker_backend_ready()
@@ -85,7 +94,7 @@ def _fake_run_command(argv: list[str], timeout_seconds: int = 15) -> ExecResult:
 
 def test_varlock_env_dir_defaults_to_managed_config_root(
     tmp_path: pathlib.Path,
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     repo_root = tmp_path / "assets"
     asset_dir = repo_root / "platform" / "configs" / "varlock"
@@ -95,7 +104,7 @@ def test_varlock_env_dir_defaults_to_managed_config_root(
     (asset_dir / ".env.ci.example").write_text("APP_ENV=ci\n", encoding="utf-8")
     (asset_dir / ".env.prod.example").write_text("APP_ENV=prod\n", encoding="utf-8")
     managed_root = tmp_path / "config-root"
-    monkeypatch.setenv("STRONGCLAW_CONFIG_DIR", str(managed_root))
+    test_context.env.set("STRONGCLAW_CONFIG_DIR", str(managed_root))
     legacy_dir = repo_root / "platform" / "configs" / "varlock"
 
     expected = runtime.varlock_env_dir(repo_root)
diff --git a/tests/suites/unit/clawops/test_strongclaw_varlock_env.py b/tests/suites/unit/clawops/test_strongclaw_varlock_env.py
index 0b9ca632..1d9ee98e 100644
--- a/tests/suites/unit/clawops/test_strongclaw_varlock_env.py
+++ b/tests/suites/unit/clawops/test_strongclaw_varlock_env.py
@@ -4,8 +4,6 @@
 
 import pathlib
 
-import pytest
-
 from clawops.strongclaw_runtime import (
     load_env_assignments,
     varlock_env_template_file,
@@ -44,10 +42,9 @@ def _varlock_validation_success(
 def test_ensure_required_defaults_generates_neo4j_credentials(
     tmp_path: pathlib.Path,
     test_context: TestContext,
-    monkeypatch: pytest.MonkeyPatch,
 ) -> None:
     local_env_file = tmp_path / ".env.local"
-    monkeypatch.setenv("VARLOCK_LOCAL_ENV_FILE", str(local_env_file))
+    test_context.env.set("VARLOCK_LOCAL_ENV_FILE", str(local_env_file))
     secrets = iter(
         (
             "gateway-secret-value",
diff --git a/tests/suites/unit/testing/test_qdrant_runtime.py b/tests/suites/unit/testing/test_qdrant_runtime.py
index 6094aa97..93ddd3ee 100644
--- a/tests/suites/unit/testing/test_qdrant_runtime.py
+++ b/tests/suites/unit/testing/test_qdrant_runtime.py
@@ -6,6 +6,7 @@
 
 import pytest
 
+from tests.plugins.infrastructure.context import TestContext
 from tests.utils.helpers.qdrant_runtime import (
     DEFAULT_QDRANT_IMAGE,
     QDRANT_IMAGE_ENV,
@@ -32,7 +33,7 @@ def test_default_qdrant_image_uses_pinned_ghcr_mirror() -> None:
 
 
 def test_require_live_url_uses_repo_pinned_image_by_default(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     """Real-mode runtimes should default to the repo-pinned Qdrant image."""
     commands: list[list[str]] = []
@@ -48,10 +49,13 @@ def _fake_run(
         commands.append(command)
         return subprocess.CompletedProcess(command, 0, stdout="container-id", stderr="")
 
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.shutil.which", _docker_path)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._reserve_local_port", lambda: 46333)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._wait_for_qdrant", _noop_wait)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.subprocess.run", _fake_run)
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.shutil.which", new=_docker_path)
+    test_context.patch.patch(
+        "tests.utils.helpers.qdrant_runtime._reserve_local_port",
+        new=lambda: 46333,
+    )
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime._wait_for_qdrant", new=_noop_wait)
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.subprocess.run", new=_fake_run)
 
     runtime = QdrantRuntime(context=None, mode="real")
     live_url = runtime.require_live_url()
@@ -63,7 +67,7 @@ def _fake_run(
 
 
 def test_require_live_url_honors_configured_image_override(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     """Real-mode runtimes should honor an explicit image override."""
     commands: list[list[str]] = []
@@ -79,11 +83,14 @@ def _fake_run(
         commands.append(command)
         return subprocess.CompletedProcess(command, 0, stdout="container-id", stderr="")
 
-    monkeypatch.setenv(QDRANT_IMAGE_ENV, "ghcr.io/example/qdrant:test")
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.shutil.which", _docker_path)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._reserve_local_port", lambda: 46333)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._wait_for_qdrant", _noop_wait)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.subprocess.run", _fake_run)
+    test_context.env.set(QDRANT_IMAGE_ENV, "ghcr.io/example/qdrant:test")
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.shutil.which", new=_docker_path)
+    test_context.patch.patch(
+        "tests.utils.helpers.qdrant_runtime._reserve_local_port",
+        new=lambda: 46333,
+    )
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime._wait_for_qdrant", new=_noop_wait)
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.subprocess.run", new=_fake_run)
 
     runtime = QdrantRuntime(context=None, mode="real")
     runtime.require_live_url()
@@ -93,7 +100,7 @@ def _fake_run(
 
 
 def test_require_live_url_skips_when_registry_access_fails(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     """Real-mode runtimes should skip when the registry is temporarily unavailable."""
 
@@ -116,9 +123,12 @@ def _fake_run(
             ),
         )
 
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.shutil.which", _docker_path)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._reserve_local_port", lambda: 46333)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.subprocess.run", _fake_run)
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.shutil.which", new=_docker_path)
+    test_context.patch.patch(
+        "tests.utils.helpers.qdrant_runtime._reserve_local_port",
+        new=lambda: 46333,
+    )
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.subprocess.run", new=_fake_run)
 
     runtime = QdrantRuntime(context=None, mode="real")
 
@@ -127,7 +137,7 @@ def _fake_run(
 
 
 def test_require_live_url_fails_on_non_registry_docker_errors(
-    monkeypatch: pytest.MonkeyPatch,
+    test_context: TestContext,
 ) -> None:
     """Real-mode runtimes should still fail for ordinary docker runtime errors."""
 
@@ -146,9 +156,12 @@ def _fake_run(
             stderr="docker: Error response from daemon: driver failed programming external connectivity",
         )
 
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.shutil.which", _docker_path)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime._reserve_local_port", lambda: 46333)
-    monkeypatch.setattr("tests.utils.helpers.qdrant_runtime.subprocess.run", _fake_run)
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.shutil.which", new=_docker_path)
+    test_context.patch.patch(
+        "tests.utils.helpers.qdrant_runtime._reserve_local_port",
+        new=lambda: 46333,
+    )
+    test_context.patch.patch("tests.utils.helpers.qdrant_runtime.subprocess.run", new=_fake_run)
 
     runtime = QdrantRuntime(context=None, mode="real")
 
diff --git a/tests/suites/unit/testing/test_test_context.py b/tests/suites/unit/testing/test_test_context.py
index 4c76de6f..aa5ecb66 100644
--- a/tests/suites/unit/testing/test_test_context.py
+++ b/tests/suites/unit/testing/test_test_context.py
@@ -3,6 +3,8 @@
 from __future__ import annotations
 
 import logging
+import os
+import pathlib
 
 import pytest
 
@@ -185,3 +187,21 @@ def test_notes_dict_available_for_diagnostics() -> None:
     ctx = TestContext()
     ctx.notes["scope"] = "diagnostic"
     assert ctx.notes["scope"] == "diagnostic"
+
+
+def test_chdir_restores_original_working_directory(tmp_path: pathlib.Path) -> None:
+    ctx = TestContext()
+    original = pathlib.Path.cwd()
+    target = tmp_path / "workspace"
+    target.mkdir()
+
+    try:
+        ctx.chdir(target)
+
+        assert pathlib.Path.cwd() == target
+
+        ctx.cleanup_all()
+
+        assert pathlib.Path.cwd() == original
+    finally:
+        os.chdir(original)
diff --git a/tests/utils/scripts/analyze_fixtures.py b/tests/utils/scripts/analyze_fixtures.py
index 50a7c2dd..0a9e1de1 100644
--- a/tests/utils/scripts/analyze_fixtures.py
+++ b/tests/utils/scripts/analyze_fixtures.py
@@ -38,7 +38,7 @@ def _test_source_files(repo_root: Path) -> list[Path]:
 def _is_direct_monkeypatch_call(node: ast.Call) -> bool:
     if not isinstance(node.func, ast.Attribute):
         return False
-    if node.func.attr not in {"delenv", "setattr", "setenv"}:
+    if node.func.attr not in {"chdir", "delenv", "setattr", "setenv"}:
         return False
     return isinstance(node.func.value, ast.Name) and node.func.value.id == "monkeypatch"