- Kimi factory: add max_tokens=16384, temperature=1.0, top_p=1.0,
and chat_template_kwargs.thinking=true for kimi models
- Add chunk count traces in stream_response so we see LLM progress
immediately in logs: 'LLM chunk #N received (len=X)'
- Keep generic stream parser clean — model-specific logic lives in
the request builder (Kimi factory pattern)
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
Previously the recv_task awaited stream_response() directly, which
froze the entire WebSocket message receiver while the LLM ran (30s+).
This meant a second user message couldn't be processed until the
first LLM call finished — a race condition that locked the session.
Now stream_response runs in its own tokio::spawn, keeping recv_task
free to handle new messages immediately. Also fixed borrow/lifetime
issue by cloning the response channel sender out of the lock scope.
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
- Move all preprocessing transforms (convert_multiword_keywords, preprocess_llm_keyword,
convert_while_wend_syntax, predeclare_variables) into BasicCompiler::preprocess_basic
so .ast files are fully preprocessed by Drive Monitor
- Replace ScriptService compile/compile_preprocessed/compile_tool_script with
single run(ast_content) that does engine.compile() + eval_ast_with_scope()
- Remove .bas fallback in tool_executor and start.bas paths - .ast only
- Remove dead code: preprocess_basic_script, normalize_variables_to_lowercase,
convert_save_for_tools, parse_save_parts, normalize_word
- Fix: USE KB 'cartas' in tool .ast now correctly converted to USE_KB('cartas')
during compilation, ensuring KB context injection works after tool execution
- Fix: add trace import in llm/mod.rs
- Log full LLM response preview (500 chars) with has_html detection
- Log WebSocket send with message type, completeness, and content preview
- Use clone() for chunk in BotResponse to ensure accurate logging
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
When a tool button like Cartas activates a KB via USE KB, instead of
returning just the tool result (empty/label), the handler now checks
if session has active KBs. If so and result is empty/trivial,
falls through to the full LLM pipeline which injects KB context.
When a bot lacks embedding-url in its own config, from_bot_config now
falls back to the default bot's config via ConfigManager::get_config.
Previously it returned empty string, causing embedding server connection
failures for bots without explicit embedding configuration.
Header detection was only checking for key,value but the actual
CSV uses name,value as header row. Now both are detected and skipped.
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
- Removed unconditional .skip(1) that was skipping first config line
- Added header detection: skips first line only if it looks like 'key,value' header
- Added validation to skip empty keys
- Also fixed indentation in drive_monitor gbkb file processing
This fixes the issue where config.csv changes on Drive weren't being
synced to bot_configuration database table for salesianos bot.
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
The bge-small-en-v1.5-f32.gguf model has n_ctx_train=512. With batch_size=16
and ~300+ tokens per chunk, total tokens exceed 512 causing GGML_ASSERT crash.
Now with batch_size=2, embeddings are processed safely.
- Remove hardcoded URL list for remote API detection
- Try /health first, then probe with HEAD if 404/405
- Re-enable embedding server ready check in drive_monitor
- No more embedding_key hack that skipped health checks entirely
- Bug 1: check_gbkb_changes now preserves indexed=true from previous
state when etag matches, preventing redundant re-indexing every cycle
- Bug 2: USE KB fallback uses bot_id_short (8 chars) instead of random
UUID, matching the collection name convention used by DriveMonitor
- Bug 3: handle_gbkb_change now upserts into kb_collections table after
successful indexing, so USE KB can find the collection at runtime
- Changed ON CONFLICT DO NOTHING to DO UPDATE for kb_collections inserts
- Changed process_gbkb_folder return type to Result<IndexingResult>
- get_work_path_default/get_stack_path no longer rely on CWD-relative botserver-stack check which caused wrong output path in production when CI left that directory
- DriveMonitor now marks .bas file states as indexed=true after list+compile cycle
- Added compile_tool logging for work_dir path
1. Fix model.starts_with('') always true - was limiting ALL models to 768 tokens
(local llama limit), truncating system prompts and KB context. Now only
applies when model=='local' or empty string, default is 32k tokens.
2. Fix create_bot_from_drive missing NOT NULL columns (llm_provider,
context_provider) - bots auto-created from S3 buckets failed to persist.
3. Fix S3 endpoint URL construction missing port 9100.
4. Fix Vault seed: vectordb.url was empty string, now defaults to
http://localhost:6333.
5. Fix Vault credential regeneration on recovery - added vault_seeds_exist().
6. Fix CA cert path for Vault TLS (botserver-stack vs botserver-stack).
7. Add bot verification after insert in create_bot_from_drive.
- Add ZITADEL_DATABASE_* environment variables to directory component env_vars
- Remove inline env vars from exec_cmd (now applied via spawn_with_envs)
- Use $DB_PASSWORD reference to fetch from Vault at runtime
- This ensures Zitadel gets database credentials on every boot, not just during install
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
