The previous fix used Handle::current().block_on() which deadlocks when
the Rhai engine runs on a Tokio worker thread — it blocks the very
thread the async task needs to make progress.
New approach: spawn a dedicated background thread with its own
single-threaded Tokio runtime, communicate via mpsc channel with a
45s timeout. This completely isolates the LLM runtime from the
caller's runtime, eliminating any possibility of thread starvation
or nested-runtime deadlock.
Also remove unused 'trace' import from llm/mod.rs.
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
Implement ADD_SWITCHER keyword following the same pattern as ADD_SUGGESTION_TOOL:
- Created switcher.rs module with add_switcher_keyword() and clear_switchers_keyword()
- Added preprocessing to convert "ADD SWITCHER" to "ADD_SWITCHER"
- Added to keyword patterns and get_all_keywords()
- Stores switcher suggestions in Redis with type "switcher" and action "switch_context"
- Supports both "ADD SWITCHER" and "ADD_SWITCHER" syntax
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
SADD stores suggestions in a set (deduplicated) instead of a list (accumulates).
get_suggestions now uses SMEMBERS instead of LRANGE. Removed the TODO about
clearing suggestions since SADD inherently prevents duplicates.
- 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
- Replace ADD SUGGESTION TOOL with ADD_SUGG_TOOL (single token)
- Replace ADD SUGGESTION TEXT with ADD_SUGG_TEXT
- Replace ADD SUGGESTION with ADD_SUGG
- Keep ADD_SUGGESTION_TOOL as legacy alias for backward compat
- Preprocessor converts ADD SUGGESTION TOOL -> ADD_SUGG_TOOL automatically
- Eliminates collision with ADD BOT, ADD MEMBER in Rhai parser
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
- 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>
- Remove LocalFileMonitor and ConfigWatcher for /opt/gbo/data
- Remove /opt/gbo/data from mount_all_bots() scanning
- Change start.bas, tables.bas, and tool paths to use work directory
- Filter drive buckets to only gbo-* prefix
- Remove unused create_bot_simple method
- Fix all warnings (unused imports, variables, dead code)
- Replace EmailService::send_email stub with full lettre SMTP implementation
- Vault resolution chain: bot-specific → default bot → system fallback
- Seed Vault prod with default email config (contato@pragmatismo.com.br)
- Update all call sites to pass bot_id for Vault lookup
- Support attachments via lettre MultiPart/Attachment API
- Remove unused imports and dead code
- Simplify constructor from (pool, bot_id, cache) to (&state, bot_id)
- Adapter now extracts conn and cache from AppState internally
- Updates 15 call sites across 6 files
- Removes redundant parameter plumbing at every call site
- Removed duplicate DbColumn struct, PROTECTED_COLUMNS const, and sync_table_schema fn
- File now has single clean implementation with column drop protection
- Add columns_dropped counter to MigrationResult
- Add PROTECTED_COLUMNS list (id, bot_id, org_id, user_id, created_at, etc.)
- Detect orphaned columns (in DB but not in tables.bas) and drop them
- Protected columns are never dropped automatically
- Uses DROP COLUMN IF EXISTS for safety
- Logs warnings for orphaned columns before dropping
- server.rs: Use PathBuf for cert_dir
- auth_routes.rs: Use PathBuf for pat_path
- qrcode.rs: Bind get_work_path() to local var before unwrap_or
- import_export.rs: Bind get_work_path() to local var in both functions (2 occurrences)
- Adds get_stack_path() helper: returns /opt/gbo in production (.env without botserver-stack), ./botserver-stack in dev
- Adds get_work_path() helper: returns /opt/gbo/work in production, ./botserver-stack/data/system/work in dev
- Updated 35+ files to use dynamic path resolution
- Production system container no longer needs botserver-stack directory
- Work files go to /opt/gbo/work instead of /opt/gbo/bin/botserver-stack
- Fixes panic: Cannot start a runtime from within a runtime
- set_user.rs was using futures::executor::block_on directly in Rhai callback
- Now uses std:🧵:spawn + new_current_thread().block_on() pattern
- This is called during bootstrap and was causing startup crash
- Fixes panic: Cannot start a runtime from within a runtime
- kb_statistics.rs: Wrap all async calls in std:🧵:spawn
- post_to.rs: Replace Handle::try_current with thread::spawn + mpsc
- Removes dead Handle::try_current checks from sync functions
- Follows AGENTS.md pattern for async-from-sync callbacks
- Fix double_ended_iterator_last: use next_back() instead of last()
- Fix manual_clamp: use .clamp() instead of min().max()
- Fix too_many_arguments: create KbInjectionContext struct
- Fix needless_borrow: remove unnecessary & reference
- Fix let_and_return: return value directly
- Fix await_holding_lock: drop guard before await
- Fix collapsible_else_if: collapse nested if-else
All changes verified with cargo clippy (0 warnings, 0 errors)
Note: Local botserver crashes with existing panic during LocalFileMonitor initialization
This panic exists in original code too, not caused by these changes
- Handle::current().block_on() panics when called from within a runtime
- replaced all 5 occurrences with std:🧵:spawn + mpsc::channel
- matches the pattern already used across other keyword files
Root cause: block_in_place + new_current_thread().block_on() panics when
called from within tokio runtime (including spawn_blocking). Tokio doesn't
allow nested block_on() calls.
Fix: Replace ALL block_in_place patterns with std:🧵:spawn + mpsc channel.
This creates a completely separate OS thread with its own runtime, avoiding
any nesting issues. Works from any context: async, spawn_blocking, or sync.
Files: 14 files across secrets, utils, state, calendar, analytics, email,
and all keyword handlers (universal_messaging, search, book, create_draft,
create_site, hearing/syntax, use_tool, find, admin_email, goals)
Root cause: Handle::current().block_on() panics inside multi_thread runtime
with 'Cannot start a runtime from within a runtime' error.
Fix: All sync-to-async bridges now use tokio::runtime::Builder::new_current_thread()
instead of Handle::current().block_on(). Also changed SECRETS_MANAGER from
tokio::sync::RwLock to std::sync::RwLock to eliminate unnecessary async overhead.
Files: 14 files across keywords, secrets, utils, state, calendar, analytics, email
Impact: Fixes production crash during bot loading phase
- Remove all std::env::var calls except VAULT_* and PORT
- get_from_env returns hardcoded defaults only (no env var reading)
- Auth config, rate limits, email, analytics, calendar all use Vault
- WORK_PATH replaced with get_work_path() helper reading from Vault
- .env on production cleaned to only VAULT_ADDR, VAULT_TOKEN, VAULT_CACERT, PORT
- All service IPs/credentials stored in Vault secret/gbo/*
