- numpy soft dependency: try/except import + _HAS_NUMPY flag; _encode_embedding
and _decode_embedding fall back to struct.pack/unpack; search_vector falls back
to pure-Python cosine loop — startup never fails without numpy reinstalled
- SQLite UPSERT guard: _HAS_UPSERT = sqlite_version_info >= (3,24,0); save_chunk
and save_chunks_batch fall back to INSERT OR REPLACE on SQLite < 3.24 with a
one-time startup warning about potential FTS rowid drift
- _bm25_rank_to_score floor: 0.3 + 0.69*(|rank|/(1+|rank|)) → always in [0.3, 0.99),
prevents small-corpus matches scoring 0.0 and being filtered by min_score
- detect_index_dim BLOB-aware: check isinstance(raw, bytes) first and return
len(raw)//4 before json.loads, so /memory status works after embedding format switch
- Comment: "CJK single-char" → "CJK tokens shorter than 3 characters"
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
matched_count only counted cjk_words hits; pure ASCII queries had
cjk_words=[] so matched_count=0 and all SQL-matched rows were filtered
out. Change to count across all tokens (cjk_words + ascii_words) so
the LIKE fallback works correctly when FTS5 is unavailable.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add embedding_provider config knob with native support for
openai / dashscope / doubao / zhipu / linkai, plus an in-chat
/memory status and /memory rebuild-index workflow for switching
vendors safely.
When reloading a conversation, failed tool calls incorrectly showed checkmark instead of X because the is_error field was lost in the history rendering pipeline. Propagate is_error from DB extraction through to the frontend rendering to match the live SSE behavior.
Further refinements on top of #2795:
- persist real session_id (notify_session_id) at task creation so group chats
correctly map back to the user's actual conversation
- mark scheduler turns with [SCHEDULED] (recognise legacy "Scheduled task"
prefix too for backward-compatible pruning)
- prune both DB and in-memory to scheduler_inject_max_per_session (default 3),
only marker-tagged pairs are touched; regular user turns never deleted
- send_message type gated by scheduler_inject_send_message (default false) —
fixed reminder text rarely benefits follow-up Q&A
Co-authored-by: huangrichao2020 <grdomai43881@gmail.com>
- New memory/deep-dream.mdx (zh/en/ja): memory flow, distillation rules, dream diary, manual trigger, safety mechanisms
- Simplify long-term memory page, link to deep-dream for details
- New cli/memory-knowledge.mdx (zh/en/ja): memory and knowledge commands
- Move knowledge commands from general.mdx to memory-knowledge.mdx
- Register new pages in docs.json navigation for all languages
- Add /memory dream to cli/index.mdx command tables
- Unified flush + context injection into a single async LLM call
(flush_from_messages accepts context_summary_callback)
- Fixed response parsing bug: handle generator returns and Claude-format
dicts from bot.call_with_tools, which previously caused all LLM
summaries to silently fail (falling back to rule-based extraction)
- Removed standalone context summary prompts and methods; reuse the
existing [DAILY]/[MEMORY] summarization pipeline
- Updated docs (zh/en/ja) to reflect the new injection behavior
- Add knowledge/ directory structure and knowledge-wiki skill for structured knowledge accumulation
- Auto-inject MEMORY.md into system prompt with truncation (last 200 lines)
- Light Dream: extend flush_memory to extract long-term memories into MEMORY.md with date stamps
- Add mandatory knowledge auto-write rules in system prompt (no user confirmation needed)
- Expand MemoryManager.sync() to index knowledge/ files for vector search
- Update RULE.md template with workspace conventions and knowledge guidelines
- Use LLM to summarize discarded context into concise daily memory entries
- Batch trim to half when exceeding max_turns/max_tokens, reducing flush frequency
- Run summarization asynchronously in background thread, no blocking on replies
- Add daily scheduled flush (23:55) as fallback for low-activity days
- Sync trimmed messages back to agent to keep context state consistent
