Knowledge files were only partitioned in the database, which made nested uploads, local folder visibility, and delete behavior diverge from the UI. This change makes folder selection drive physical storage paths, keeps original filenames, adds a minimal WebDAV mount/sync path, and reshapes the knowledge panel so local and remote sources can share the same surface.
Constraint: Existing knowledge flow already depends on local-folder-backed uploads and document indexing
Rejected: Real-time bidirectional WebDAV sync | too much conflict and lifecycle complexity for the first pass
Confidence: medium
Scope-risk: moderate
Reversibility: messy
Directive: Keep remote mounts single-direction into local knowledge folders until etag-based incremental sync and conflict rules are verified
Tested: Python py_compile on new/modified backend files; LSP diagnostics on new frontend/backend files; manual targeted code-path inspection
Not-tested: Full pytest/vitest end-to-end runs blocked by environment temp/cache permission errors; live WebDAV server interoperability
- M.2: ForgettingCurve, MemoryDecay, MemoryReinforcement (selective forgetting)
- M.3: DailyDigestGenerator, ReminderScheduler, ProactiveInformer (proactive reminders)
- M.4: MemoryExtractor with LLM-based memory extraction from conversations
- M.5: MemoryRecallInjector with token budget control for prompt injection
- All phases include comprehensive unit tests (109 tests passing)
- Updated checklist.md to mark all tasks complete
Add Day 1 agent runtime foundations with task and event schemas, verifier support, capability metadata, graph event tracing, and regression coverage while preserving the direct execution path.
Align the L3 graph, agent service, and sync tool shims on one canonical continuity contract so clarification resumes and persisted snapshots behave consistently. Add targeted regressions and hardening notes covering system-message coalescing, async bridge usage, and continuity rehydration.
Persist streaming chat state during generator cleanup, close the SSE inner stream safely, and reject cross-user conversation access while locking the behavior with focused regressions.
Make the project start more reliably in the current Windows bash setup, add a safe root .env.example for onboarding, and lower the backend Python floor to 3.11 to match the validated local environment.
Add sub-commander orchestration updates, align frontend integrations, and refine knowledge view behavior without including local data artifacts.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Introduce the backend pieces for brain memory ingestion, routing, and
system telemetry so the new knowledge workflows can project data into a
brain view. The supporting tests lock in the new behavior and keep the
expanded backend surface stable.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Normalize uploaded documents into structured markdown, add clearer parser
errors for missing dependencies, and cover the ingestion flow with
backend tests. This also replaces deprecated UTC timestamp helpers in
the touched backend paths so the knowledge pipeline stays warning-free.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Make runtime log queries support request correlation and date-range diagnostics with shared filtering semantics so the log page can use one consistent contract.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
