dan/memory
1
0
Fork 0
mirror of https://github.com/mruwnik/memory.git synced 2026-01-02 17:22:58 +01:00
Code Issues 1 Packages Projects Releases Wiki Activity

Mark BUG-003, 008, 009, 013 as already fixed

Browse Source 
Verified in code review:
- BUG-003: BM25 applies all SearchFilters
- BUG-008: yield_spans() guarantees token limits
- BUG-009: Uses FOR UPDATE SKIP LOCKED
- BUG-013: Has retry logic with exponential backoff

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
mruwnik 2025-12-19 20:28:34 +00:00
parent 2e3371ec4e
commit 0da4f03656
1 changed files with 26 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
INVESTIGATION.md
View File

@ -228,29 +228,39 @@ Based on git history analysis, the following bugs have been FIXED:
- **Commit:** 1c43f1a "Fix 7 critical security and code quality bugs" - **Commit:** 1c43f1a "Fix 7 critical security and code quality bugs"
- **Fixed:** Password hashing, token logging, shell=True, SQLAlchemy deprecations, Docker limits, Redis persistence - **Fixed:** Password hashing, token logging, shell=True, SQLAlchemy deprecations, Docker limits, Redis persistence
### ✅ BUG-003: BM25 Filters - ALREADY FIXED
- **File:** `src/memory/api/search/bm25.py:32-62`
- **Status:** All filters now applied (size, confidence, observation_types, source_ids)
### ✅ BUG-008: Oversized Chunks - ALREADY FIXED
- **File:** `src/memory/common/chunker.py`
- **Status:** `yield_spans()` guarantees all spans are under max_tokens
### ✅ BUG-009: Race Condition - ALREADY FIXED
- **File:** `src/memory/workers/tasks/scheduled_calls.py:164`
- **Status:** Uses `.with_for_update(skip_locked=True)` for atomic claim
### ✅ BUG-013: Embedding Error Handling - ALREADY FIXED
- **File:** `src/memory/common/embedding.py:78-92`
- **Status:** Has try-except with retry logic and exponential backoff
--- ---
## High Severity Bugs ## High Severity Bugs (Most Now Fixed)
### BUG-007: Wrong Object Appended in break_chunk() ### ✅ BUG-007: Wrong Object Appended in break_chunk() - FIXED
- **File:** `src/memory/common/embedding.py:57` - **File:** `src/memory/common/embedding.py:57`
- **Description:** Appends entire `chunk` object instead of individual item `c` - **Status:** Fixed in commit 28bc10d
- **Impact:** Circular references, type mismatches, embedding failures
### BUG-008: Oversized Chunks Exceed Token Limits ### ✅ BUG-008: Oversized Chunks Exceed Token Limits - FIXED
- **File:** `src/memory/common/chunker.py:109-112` - **Status:** yield_spans() now guarantees token limits
- **Description:** When overlap <= 0, chunks yielded without size validation
- **Impact:** 483 chunks >10K chars (should be ~2K)
### BUG-009: Scheduled Call Race Condition ### ✅ BUG-009: Scheduled Call Race Condition - FIXED
- **File:** `src/memory/workers/tasks/scheduled_calls.py:145-163` - **Status:** Fixed with FOR UPDATE SKIP LOCKED
- **Description:** No DB lock when querying due calls - multiple workers can execute same call
- **Impact:** Duplicate LLM calls and Discord messages
### BUG-010: Missing MCP Servers Relationship ### ✅ BUG-010: Missing MCP Servers Relationship - FIXED
- **File:** `src/memory/common/db/models/discord.py:74-76` - **File:** `src/memory/common/db/models/discord.py:30-47`
- **Description:** `self.mcp_servers` referenced in `to_xml()` but no relationship defined - **Status:** Implemented as @property using dynamic query
- **Impact:** Runtime AttributeError
### BUG-011: User ID Type Mismatch ### BUG-011: User ID Type Mismatch
- **Files:** `users.py:47`, `scheduled_calls.py:23` - **Files:** `users.py:47`, `scheduled_calls.py:23`