🔐 Add user-level access control via OIDC#153
Open
StephanMeijer wants to merge 1 commit into
Open
Conversation
StephanMeijer
force-pushed
the
feat/bring-everything-up
branch
3 times, most recently
from
26cac48 to
e7ac7ce
Compare
There was a problem hiding this comment.
Pull request overview
This PR introduces user-level access control for the search endpoint by supporting both service-token and OIDC authentication, extracting the authenticated user subject for filtering, and documenting the query DSL and access-control model.
Changes:
- Added Bolt route-level auth middleware to support service token auth, OIDC auth, and a combined “search auth” mode.
- Updated search query construction (explicit language fields + updated trigram settings) and adjusted handlers to build responses from localized
_sourcefields. - Added/updated operational config and documentation (django-bolt integration, pytest-django DB config, and expanded query schema docs), plus regenerated VCR cassettes.
Reviewed changes
Copilot reviewed 18 out of 19 changed files in this pull request and generated 5 comments.
Show a summary per file
| File | Description |
|---|---|
| src/backend/uv.lock | Adds uvicorn[standard] and related transitive deps to the lockfile (dev extra). |
| src/backend/pyproject.toml | Adds uvicorn[standard] to dev dependencies. |
| src/backend/find/settings.py | Registers django_bolt, configures BOLT_API, adds a DB config, and adjusts trigram defaults. |
| src/backend/find/asgi.py | Switches ASGI entrypoint import to core.api. |
| src/backend/demo/tests/cassettes/test_commands_create_demo.yaml | Regenerates demo cassette to match updated indexing/query behavior. |
| src/backend/core/tests/conftest.py | Updates test fixtures to patch new middleware-based auth paths. |
| src/backend/core/tests/cassettes/TestSearchDocumentsHandler.test_search_with_where_clause_returns_200.yaml | Updates cassette to match new query format and _source selection. |
| src/backend/core/tests/cassettes/TestSearchDocumentsHandler.test_search_with_sort_and_limit_returns_200.yaml | Updates cassette to match new query format and _source selection. |
| src/backend/core/tests/cassettes/TestSearchDocumentsHandler.test_search_simple_query_returns_200.yaml | Updates cassette to match new query format and _source selection. |
| src/backend/core/services/search.py | Adds logging and revises full-text query field selection (explicit language fields + trigram query structure). |
| src/backend/core/middleware.py | Introduces ServiceAuthMiddleware, OIDCAuthMiddleware, and SearchAuthMiddleware. |
| src/backend/core/handlers.py | Migrates endpoints to middleware-driven auth and updates result mapping for localized fields. |
| src/backend/core/enums.py | Adjusts _source field selection to use title.* / content.*. |
| src/backend/core/bolt_auth.py | Removes service-token auth backend/guards and narrows module scope to OIDC auth. |
| src/backend/core/authentication.py | Reworks OIDC resource-server backend init to avoid requiring Django auth user model. |
| src/backend/core/api.py | Adds a django-bolt autodiscovery entrypoint that re-exports handlers.api. |
| docs/SEARCH_QUERY_SCHEMA.md | Adds comprehensive documentation for the search query schema, operators, fields, and access control. |
| docs/architecture.md | Documents system-scope vs user-scope filtering and the dual-auth search flow. |
| compose.yml | Runs the backend via manage.py runbolt in docker-compose for development. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
StephanMeijer
commented
May 16, 2026
StephanMeijer
force-pushed
the
feat/bring-everything-up
branch
from
e7ac7ce to
fa603bc
Compare
Signed-off-by: Stephan Meijer <me@stephanmeijer.com>
StephanMeijer
force-pushed
the
feat/bring-everything-up
branch
from
fa603bc to
13e1f92
Compare
Comment on lines
+74
to
+75
| must=[multi_match_standard], | ||
| should=[multi_match_trigram], |
Comment on lines
200
to
208
| # Trigrams search settings | ||
| TRIGRAMS_BOOST = values.Value( | ||
| default=0.25, environ_name="TRIGRAMS_BOOST", environ_prefix=None | ||
| default=0.01, environ_name="TRIGRAMS_BOOST", environ_prefix=None | ||
| ) | ||
| TRIGRAMS_MINIMUM_SHOULD_MATCH = values.Value( | ||
| default="75%", environ_name="TRIGRAMS_MINIMUM_SHOULD_MATCH", environ_prefix=None | ||
| default="100%", | ||
| environ_name="TRIGRAMS_MINIMUM_SHOULD_MATCH", | ||
| environ_prefix=None, | ||
| ) |
| ] | ||
|
|
||
| # Django-Bolt API autodiscovery | ||
| BOLT_API = ["core.handlers:api"] |
Comment on lines
+245
to
+251
| These fields can be used in `where` clauses by API clients: | ||
|
|
||
| | Field | Type | Description | | ||
| |-------|------|-------------| | ||
| | `id` | `string` | Document UUID (mapped to `_id` in OpenSearch) | | ||
| | `title` | `string` | Document title | | ||
| | `content` | `string` | Document content | |
Comment on lines
+261
to
+270
| ### System Fields (Server-Side Only) | ||
|
|
||
| These fields are used internally for access control: | ||
|
|
||
| | Field | Type | Description | | ||
| |-------|------|-------------| | ||
| | `is_active` | `bool` | Document is active (not deleted) | | ||
| | `users` | `list[string]` | User subs with explicit access | | ||
| | `groups` | `list[string]` | Group slugs with access | | ||
| | `service` | `string` | Service that owns the document | |
Comment on lines
101
to
+108
| # ---- Production image ---- | ||
| FROM core AS backend-production | ||
|
|
||
| # Gunicorn | ||
| RUN mkdir -p /usr/local/etc/gunicorn | ||
| COPY docker/files/usr/local/etc/gunicorn/find.py /usr/local/etc/gunicorn/find.py | ||
|
|
||
| # Un-privileged user running the application | ||
| ARG DOCKER_USER | ||
| USER ${DOCKER_USER} | ||
|
|
||
| # The default command runs gunicorn WSGI server in find's main module | ||
| CMD ["gunicorn", "-c", "/usr/local/etc/gunicorn/find.py", "find.wsgi:application"] | ||
| CMD ["python", "manage.py", "runbolt", "--host", "0.0.0.0", "--port", "8000"] |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
SearchAuthMiddlewaresupporting both service tokens and OIDC tokensuser_subfrom authenticated OIDC user for access controlwhereclause for user-level filtering (reach != restricted OR user_sub IN users OR id IN visited)docs/SEARCH_QUERY_SCHEMA.mdChanges
core/middleware.py: NewSearchAuthMiddlewarewith dual auth supportcore/services/search.py:combine_with_system_scope()for where clause mergingcore/handlers.py: Extract user fromrequest.state["user"].subfind/settings.py: Add dummy DATABASES for pytest-django compatibilityTesting
core/apps.pyandcore/bolt_auth.py(not from this PR)