- Added output for commit SHA in the Docker build workflow to facilitate tagging.
- Introduced a new verification step to ensure both architectures are present for each required variant before proceeding with the release.
- Updated the finalize_release job to tag the correct commit after successful builds and verifications.
- Introduced the ZERO_AUTO_RESET environment variable to enable automatic reset of the SQLite replica in case of replication halts.
- Updated Docker Compose files to include ZERO_AUTO_RESET in service configurations.
- Enhanced documentation to clarify the purpose and usage of the new variable.
- Introduced a new docker-compose.gpu.yml file to define GPU resource reservations for backend services.
- Updated .env.example to include GPU-related environment variables and usage instructions.
- Enhanced install.ps1 and install.sh scripts to support GPU variant selection and validation for GPU count.
- Improved error handling and user feedback for invalid GPU configurations.
- Added support for multiple image variants (CPU, CUDA 12.8, CUDA 12.6) in the Docker build workflow.
- Updated Docker Compose configuration to utilize the new SURFSENSE_VARIANT environment variable for selecting image variants.
- Enhanced documentation to clarify usage of backend image variants and their corresponding environment variable settings.
- Introduced a new optional dependency group for CUDA 12.6 in pyproject.toml.
- Updated conflict resolution to include CUDA 12.6 alongside existing configurations.
- Added source configurations for PyTorch and torchvision for CUDA 12.6.
The User mapper eager-loads the oauth_accounts collection via joined load
under AUTH_TYPE=GOOGLE, so the mint endpoint's query must call .unique()
before scalar_one_or_none() to avoid InvalidRequestError (500).
- Updated Dockerfile to conditionally install PyTorch with CPU or CUDA support based on build arguments.
- Added optional dependencies for CPU and CUDA versions of PyTorch in pyproject.toml.
- Configured uv.lock to manage dependencies and conflicts between CPU and CUDA installations.
Recursive pass over the agents module to make docstrings and inline
comments concise and intent-oriented: drop narration that just restates
the code, condense verbose module/function docstrings, and keep only the
non-obvious "why" notes. No functional code changed.
When app.notifications is the import entry point (e.g. Celery loading
app.notifications.service before any ORM code), app.db re-entered the
half-initialized app.notifications.persistence at its model-registration import
and failed with "cannot import name 'Notification' ... partially initialized".
Import app.db at the top of app/notifications/__init__.py so db fully initializes
(including its own Notification registration) before we re-import from
.persistence. Pre-existing issue (reproduces on the base commit); surfaced by the
celery worker startup.
The knowledge_base subagent imported subagent_invoke_config + EXCLUDED_STATE_KEYS
from main_agent's checkpointed_subagent_middleware -- a subagent reaching into
main-agent internals. Both symbols (plus the recursion-limit constant they need)
are a subagent-invocation contract shared by the orchestrator's task middleware
and any nested-invoking subagent. Move them to subagents/shared/invocation.py;
config.py keeps the HITL resume side-channel and constants.py keeps the
main-agent tuning knobs. All consumers (task_tool, kb tool, tests) repointed.
build_credentials/get_token_encryption are Google-OAuth helpers used by both the
Gmail and Calendar connector tools. They lived inside gmail/tools/_helpers.py,
forcing calendar -> gmail coupling. Move them to a neutral connector-level module
(connectors/google_auth.py); gmail/_helpers.py re-exports them under the legacy
private names so existing gmail tools are untouched, and calendar now imports the
shared module directly.
The KB-persistence impl lived in shared/middleware/ but no subagent uses it --
consumers are the main_agent builder and the boundary event loop. Colocate with
its owner using the folder-per-middleware shape; __init__ re-exports the public
surface. Tests that reached module internals now alias the .middleware submodule.
main_agent/middleware/kb_persistence.py -> kb_persistence/builder.py
shared/middleware/kb_persistence.py -> kb_persistence/middleware.py
The busy-mutex impl (BusyMutexMiddleware + cancel/turn-lifecycle primitives)
lived in shared/middleware/ but no subagent uses it -- consumers are the
main_agent builder and the boundary (turn lifecycle). Colocate with its owner
using the folder-per-middleware shape; __init__ re-exports the public surface so
boundary import sites only change package path:
main_agent/middleware/busy_mutex.py -> busy_mutex/builder.py
shared/middleware/busy_mutex.py -> busy_mutex/middleware.py
memory (builder) + memory_injection (impl) lived in shared/middleware/ but are
consumed only by main_agent (no subagent, no shared plumbing). Colocate with
their owner using the folder-per-middleware shape:
shared/middleware/memory.py -> main_agent/middleware/memory/builder.py
shared/middleware/memory_injection.py -> main_agent/middleware/memory/middleware.py
Re-decide subagent_stack placement using the primary-built-for lens rather
than consumer-only: it assembles the middleware stack threaded into every
subagent, so its domain is subagents -- even though main_agent is its sole
caller (analogous to subagents/registry.py, also invoked from main_agent).
Since no subagent *sibling* imports it, it does not belong in subagents/shared/
but at the subagents/ package root:
main_agent/middleware/subagent_stack.py
-> subagents/middleware_stack.py
subagents/shared/middleware/ held build_subagent_middleware_stack, but no
subagent package imports it -- its only caller is main_agent/middleware/stack.py
(the generic pack_subagent builder merely consumes the resulting dict at
runtime). It is main_agent's policy for which middleware to thread into
subagents, so it belongs with its caller:
subagents/shared/middleware/middleware_stack.py
-> main_agent/middleware/subagent_stack.py
subagents/shared/ now holds only genuinely subagent-shared code (md_file_reader,
snippets, spec, subagent_builder, hitl).
Per-file verification of the slice-3 candidates showed receipts/ and
date_filters.py are shared contracts (consumed by shared/state + shared
middleware + subagents), so they correctly stay put.
permissions was the real misfit: the rule *model* lived at shared/permissions.py
while its enforcement lived at shared/middleware/permissions/. Unify them into a
single self-contained subsystem:
shared/permissions.py -> shared/permissions/model.py
shared/middleware/permissions/{deny,ask,middleware}
-> shared/permissions/{deny,ask,middleware}
The package __init__ re-exports the model API + build_permission_mw, so the 32
external model consumers keep importing `from ...shared.permissions import Rule`
unchanged; only the 8 internal files redirect to `.model` (cycle-safe, model
loaded before middleware).
Move the lower-level runtime/infra modules out of multi_agent_chat/shared/
(they were never used by subagents, so they failed the shared-by-all-siblings
rule) and unify them with the already-relocated checkpointer:
agents/runtime/ -> agents/chat/runtime/
mac/shared/errors.py -> chat/runtime/errors.py
mac/shared/llm_config.py -> chat/runtime/llm_config.py
mac/shared/prompt_caching.py -> chat/runtime/prompt_caching.py
mac/shared/mention_resolver.py -> chat/runtime/mention_resolver.py
mac/shared/path_resolver.py -> chat/runtime/path_resolver.py
These sit below the agent packages: the boundary + agent factory + shared
middleware depend on them, and they import no agent code (acyclic).
shared/sandbox.py was used only by the filesystem middleware/tools (and the
boundary) -- never by main_agent or subagents as shared code. Move it next to
its only agent-side consumer:
multi_agent_chat/shared/sandbox.py
-> multi_agent_chat/shared/middleware/filesystem/sandbox.py
Recursive shared-folder rule: a shared/ must be shared by ALL siblings at its
level. The kernel (context, compaction, retry_after, web_search) was shared by
only 2 of the agents -- anonymous_chat + multi_agent_chat -- never by podcaster
or video_presentation. Those 2 are the "chat" category, so their shared code
belongs in that category's shared/, not the top-level one.
app/agents/anonymous_chat/ -> app/agents/chat/anonymous_chat/
app/agents/multi_agent_chat/ -> app/agents/chat/multi_agent_chat/
app/agents/shared/ -> app/agents/chat/shared/ (anon<->mac kernel)
Top-level app/agents/shared/ is gone: nothing was shared across all three
categories (chat / podcaster / video_presentation).
~289 import sites rewritten (app.agents.{anonymous_chat,multi_agent_chat,shared}
-> app.agents.chat.*); all moves are git renames (history preserved).
app/agents/ now: chat/, podcaster/, video_presentation/, runtime/.
These were never shared with anonymous_chat (nor podcaster/video_presentation)
-- only multi_agent_chat (subagents/main agent) and the boundary use them:
shared/tools/mcp/ -> multi_agent_chat/shared/tools/mcp/
shared/tools/hitl.py -> multi_agent_chat/shared/tools/hitl.py
shared/tools/catalog.py -> multi_agent_chat/shared/tools/catalog.py
shared/middleware/dedup_tool_calls.py
-> multi_agent_chat/shared/middleware/dedup_tool_calls.py
app/agents/shared/ now holds only the genuine anon<->mac kernel:
context, middleware/{compaction,retry_after}, tools/web_search.
Neither module is imported by any sibling agent package, so neither belongs in
the cross-agent shared kernel:
- checkpointer.py -> app/agents/runtime/checkpointer.py
LangGraph Postgres checkpoint saver. It's cross-agent *runtime infra* wired by
the boundary (app lifespan + anonymous_chat & multi_agent_chat flows), not
agent code. New app/agents/runtime/ layer holds boundary-wired agent infra.
- shared/system_prompt.py + shared/prompts/ -> app/prompts/
The legacy single-agent prompt composer. The live agents don't use it
(main_agent has its own system_prompt/ builder; anonymous_chat builds inline);
its only consumer is new_llm_config_routes for displaying default instructions.
Moved to the existing non-agent prompt domain:
system_prompt.py -> app/prompts/default_system_instructions.py
prompts/ -> app/prompts/system_prompt_composer/
app/agents/shared/ now contains only genuinely cross-agent code: context,
middleware/{compaction,retry_after,dedup_tool_calls}, tools/.
NOTE: get_default_system_instructions() (LLM-config UI) composes from the legacy
library, which differs from what the live agents actually run -- pre-existing
latent staleness, not changed here.
app/agents/shared/ is a sibling of anonymous_chat/podcaster/multi_agent_chat/
video_presentation, so it should only hold code shared across 2+ of those
agents. In practice podcaster and video_presentation import nothing from it,
and anonymous_chat needs only context + compaction + retry_after + web_search.
Everything else was multi_agent_chat-only (the boundary just passes through).
Move the multi_agent_chat-only cluster into multi_agent_chat/shared/ (files
moved verbatim via git rename; ~116 import sites rewritten):
errors, feature_flags, filesystem_selection, path_resolver, prompt_caching,
sandbox, llm_config, mention_resolver
middleware/busy_mutex, middleware/kb_persistence
busy_mutex/llm_config/mention_resolver are boundary-only but import the moved
modules, so they were folded in to avoid a backwards shared -> multi_agent_chat
dependency. main_agent builders now import the impls directly; the shared
middleware barrel keeps only the genuinely-shared compaction + retry_after.
Also delete the dead leftover shared/plugins and shared/skills dirs (live
copies already live under main_agent/).
Remaining in app/agents/shared/: context, system_prompt(+prompts), checkpointer,
middleware/{compaction,retry_after,dedup_tool_calls}, tools/. checkpointer and
system_prompt are boundary-only infra pending a dedicated home decision.
app/agents/shared/middleware/permission.py was an older, monolithic
PermissionMiddleware superseded by the modular permissions/ package under
multi_agent_chat/shared/middleware/ (core + evaluation + ask/ + factory).
Production wires only the package (main_agent stack + every subagent
builder); the kernel file was reachable only through the shared barrel
re-export (itself unused) and two tests pinned to its dead internals
(_raise_interrupt, _normalize_permission_decision, old after_model shape).
- delete app/agents/shared/middleware/permission.py
- drop PermissionMiddleware from the shared middleware barrel
- delete test_permission_middleware.py (covered the dead impl only; live
behavior is covered by tests/.../middleware/shared/permissions/*)
- test_desktop_safety_rules.py: keep the ruleset-level regression tests,
drop the dead import + TestPermissionMiddlewareIntegration class
knowledge_search, memory_injection and scoped_model_fallback no longer
belong in the cross-agent kernel (app/agents/shared/middleware): they are
consumed only inside multi_agent_chat. Relocate each impl next to the
builder that uses it:
- knowledge_search.py -> multi_agent_chat/shared/middleware/ (genuinely
shared: its _render_priority_message feeds kb_context_projection, used by
both the main agent and the KB subagent)
- memory_injection.py -> multi_agent_chat/shared/middleware/ (beside its
memory.py builder)
- scoped_model_fallback.py -> multi_agent_chat/shared/middleware/resilience/
(beside fallback.py/bundle.py)
Impls moved verbatim (git rename). Builders/consumers now import the local
sibling; main_agent knowledge_priority imports the new shared path; shared
middleware barrel trimmed.
Tests: repoint imports; convert the knowledge_search monkeypatch targets
from brittle dotted-string form to object-based patching (monkeypatch.setattr
on the imported module), which is robust to import ordering. No behavior
change.
Each main-agent-only middleware now lives in its own folder under
main_agent/middleware/<concept>/ with builder.py (flag-gated construction)
+ middleware.py (the impl), re-exported via __init__.py. This kills the
cross-folder hop into agents/shared/middleware and keeps each middleware's
two responsibilities (build vs behavior) as colocated siblings.
Moved (impl from shared/middleware, builder from main_agent/middleware):
action_log, anonymous_document, context_editing, doom_loop, knowledge_tree,
noop_injection, otel_span, tool_call_repair.
Impls moved verbatim (git rename, no body edits) so behavior is unchanged.
Builders now import from the local .middleware sibling. stack.py import
paths updated for the 3 renamed folders; shared middleware barrel trimmed;
tests repointed (imports + patch targets).
DedupHITLToolCallsMiddleware is only wired by the main_agent stack, but
its module also exports dedup-key resolvers consumed by the shared MCP
tool layer. Splitting keeps the resolvers (dedup_key_full_args,
wrap_dedup_key_by_arg_name, DedupResolver) in shared and moves the
middleware class verbatim into main_agent/middleware/dedup_hitl.py
(merged with its builder), eliminating the shared->main_agent dependency
that a flat move would create. No behavior change.
file_intent (FileIntentMiddleware) and flatten_system
(FlattenSystemMessageMiddleware) were only ever instantiated in the
single-agent chat_deepagent stack, which was removed in 14bbea085. They
have no production consumer in multi_agent_chat. Delete both modules and
their unit tests.
Also drop the vestigial KnowledgeBaseSearchMiddleware alias (= the live
KnowledgePriorityMiddleware); its tests now target the real class so the
behavior coverage is preserved. Trim the three barrel/__all__ entries and
strip the now-dead class names from comments.
The concrete filesystem backends are consumed only by the MAC filesystem
layer (tools, path-resolution middleware, the resolver, skills backend) and
tests -- no external app code. Group them next to the filesystem middleware
they serve:
- filesystem_backends.py -> filesystem/backends/resolver.py
- middleware/kb_postgres_backend.py -> filesystem/backends/kb_postgres.py
- middleware/local_folder_backend.py -> filesystem/backends/local_folder.py
- middleware/multi_root_local_folder_backend.py -> .../multi_root_local_folder.py
- document_xml.py -> filesystem/backends/document_xml.py
Repoint all 21 importers. No behavior change; import-all + filesystem
backend/path-resolution/knowledge-search unit tests stay green (478).
permissions.py (authorization Rule/Ruleset model) is consumed across all
MAC subagents + the permissions middleware, with a single external
consumer (user_tool_allowlist service) -> move to
multi_agent_chat/shared/permissions.py and repoint all 42 sites.
deliverable_wait.py (wait_for_deliverable) is used only by the podcast and
video_presentation deliverable tools -> colocate into
subagents/builtins/deliverables/.
No behavior change; import-all + permission/allowlist/deliverable unit
tests stay green.
receipt.py (Receipt model + make_receipt) and receipt_command.py
(with_receipt Command helper) are a tight pair used only by MAC subagent
tools, the graph state, and the kb_persistence middleware -- no external
code imports them (the streaming tool_end handler only references them in a
docstring). Move both into a dedicated receipts/ package
(receipts/receipt.py + receipts/command.py) and repoint importers.
No behavior change; import-all + receipt/deliverable unit tests stay green.
filesystem_state.py (the multi-agent graph state) and state_reducers.py
(its merge reducers) are consumed only by multi_agent_chat (filesystem
tools/middleware, kb projection, and the MAC-only shared middleware) plus
two unit tests -- no external app code. Relocate them into a dedicated
multi_agent_chat/shared/state/ package (filesystem_state.py + reducers.py)
and repoint every importer.
No behavior change; import-all + the full unit/middleware + unit/agents
suites (1066 tests) stay green.
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.
