shared/tools/knowledge_base.py had exactly one production consumer: the
report deliverable, which imported it via `from .knowledge_base import ...`
-- a sibling path that did not exist, so the report KB-search path would
raise ImportError at runtime.
Move the module next to report.py (subagents/builtins/deliverables/tools/)
which makes that relative import valid, and move its only dependency
(shared/utils.py date helpers) to multi_agent_chat/shared/date_filters.py,
shared between the KB tool and the knowledge_search middleware.
Drop the now-unused knowledge-base re-exports from the shared/tools barrel
and repoint the integration tests. import-all + error-contract stay green.
Eliminate the top-level multi_agent_chat/middleware/ package so each slice
owns its middleware (vertical-slice colocation):
- middleware/shared/ -> shared/middleware/ (cross-slice middleware)
- middleware/subagent/ -> subagents/shared/middleware/ (subagent stack)
- main_agent/middleware/ already colocated in Slice A
The moved shared/ subtree is internally consistent (all relative imports
stay within it), so only external absolute refs were rewritten. The
subagent stack's ..shared.* relatives were promoted to absolute paths to
the new shared/middleware/ location.
multi_agent_chat/ root is now: main_agent/, shared/, subagents/.
Verified: 2430 unit tests pass, 1 skipped (baseline unchanged).
The single-agent-era filesystem middleware (app/agents/shared/middleware/
filesystem.py, ~2000 lines) was never instantiated in production, yet three
unit suites validated it — an illusory guardrail while the live decomposed
middleware (multi_agent_chat/middleware/shared/filesystem) was unguarded.
Close the gap before reorganizing the agents module:
- Add 14 integration tests driving live B's tools in desktop mode (real
on-disk effects) and cloud mode (in-state staging, namespace policy).
- Port all high-value dead-twin assertions onto the live path: cloud rm/rmdir
staging + guard rails, KBPostgresBackend delete-view filter, mode-scoped
system prompt, cwd/relative/namespace resolution, multi-root mount
normalization.
- Delete dead twin filesystem.py, drop its __init__ re-export, and retire its
3 dead-twin tests.
Verified: test_import_all + middleware unit + FS integration all green.
Relocate the entire new_chat/tools/ package (62 files incl. registry, hitl, MCP
cluster, and all connector subpackages: gmail/slack/discord/teams/drive/etc.)
to the shared kernel. The package turned out to be a clean cohesive cluster:
its only references to non-tools new_chat modules were comments, and its
middleware deps were already flipped to shared in slice 5c.
Flip 33 live importers (multi-agent, flows, routes, services, anonymous_agent,
tests). Re-export shims remain for the frozen single-agent stack: a package
__init__ mirroring the public surface (new_chat.__init__ imports it) plus
invalid_tool + registry submodule shims (chat_deepagent imports those).
Resolves slice 5c's two transient back-edges: shared/middleware/action_log
(TYPE_CHECKING ToolDefinition) and tool_call_repair (local INVALID_TOOL_NAME)
now point at app.agents.shared.tools.
Relocate the entire new_chat/middleware/ package to the shared kernel as one
cohesive unit (it is live shared infrastructure: the multi-agent stack wraps
nearly every middleware via multi_agent_chat/middleware/main_agent/*, and
anonymous_agent consumes it too). Flip 69 live importers across both the
package-path and submodule-path forms.
Shims left for the frozen single-agent stack: a package __init__ re-export plus
submodule shims for permission, skills_backends, and scoped_model_fallback
(the three imported via submodule path by chat_deepagent/subagents).
Cycle break: importing shared.middleware previously reached back into
new_chat.tools at module load, which dragged in new_chat.__init__ ->
chat_deepagent -> the middleware shim -> half-initialized shared.middleware.
Made action_log's ToolDefinition import TYPE_CHECKING-only and
tool_call_repair's INVALID_TOOL_NAME import function-local. These tools-package
back-edges fully resolve in slice 6.
Asset note: skills_backends._default_builtin_root now walks to
app/agents/new_chat/skills/builtin (the skills/ tree migrates in slice 7).
- Included server_time_utc in the connect response schema for better synchronization.
- Updated obsidian_connect function to set server_time_utc during connection handling.
- Enhanced integration tests to verify the presence of server_time_utc in responses.
- Improved connectivity status recovery in the sync engine for better error management.
- Added a new Celery task for indexing non-markdown attachments.
- Enhanced the Obsidian plugin schema to support binary attachments.
- Updated routes to enqueue binary attachments for background processing.
- Improved metadata handling for binary attachments during indexing.
- Added tests for binary attachment processing and validation.
- Added functionality to create and update notifications during the Obsidian sync process.
- Improved handling of sync completion and failure notifications.
- Updated connector naming convention in various locations for consistency.
Build and Push Docker Images / tag_release (push) Waiting to run
Build and Push Docker Images / build (./surfsense_backend, ./surfsense_backend/Dockerfile, backend, surfsense-backend, ubuntu-24.04-arm, linux/arm64, arm64) (push) Blocked by required conditions
Build and Push Docker Images / build (./surfsense_backend, ./surfsense_backend/Dockerfile, backend, surfsense-backend, ubuntu-latest, linux/amd64, amd64) (push) Blocked by required conditions
Build and Push Docker Images / build (./surfsense_web, ./surfsense_web/Dockerfile, web, surfsense-web, ubuntu-24.04-arm, linux/arm64, arm64) (push) Blocked by required conditions
Build and Push Docker Images / build (./surfsense_web, ./surfsense_web/Dockerfile, web, surfsense-web, ubuntu-latest, linux/amd64, amd64) (push) Blocked by required conditions
Build and Push Docker Images / create_manifest (backend, surfsense-backend) (push) Blocked by required conditions
Build and Push Docker Images / create_manifest (web, surfsense-web) (push) Blocked by required conditions
- Introduced a `ProcessingMode` enum to differentiate between basic and premium processing modes.
- Updated `EtlRequest` to include a `processing_mode` field, defaulting to basic.
- Enhanced ETL pipeline services to utilize the selected processing mode for Azure Document Intelligence and LlamaCloud parsing.
- Modified various routes and services to handle processing mode, affecting document upload and indexing tasks.
- Improved error handling and logging to include processing mode details.
- Added tests to validate processing mode functionality and its impact on ETL operations.
- Added a static method `estimate_pages_from_metadata` to `PageLimitService` for estimating page counts based on file metadata.
- Integrated page limit checks in Google Drive, Dropbox, and OneDrive indexers to prevent exceeding user quotas during file indexing.
- Updated relevant indexing methods to utilize the new page estimation logic and enforce limits accordingly.
- Enhanced tests for page limit functionality, ensuring accurate estimation and enforcement across different file types.
- Updated maximum file size limit to 500 MB per file.
- Removed restrictions on the number of files per upload and total upload size.
- Enhanced handling of user-mentioning documents in the knowledge base search middleware.
- Improved document reading and processing logic to accommodate new features and optimizations.
