agentdb 1.3.0 → 1.3.1

package/README.md

@@ -359,7 +359,7 @@ npm install agentdb
 claude mcp add agentdb npx agentdb@latest mcp start
 ```
 
-This automatically configures Claude Code with all 14 AgentDB tools.
+This automatically configures Claude Code with all 29 AgentDB tools.
 
 **Manual Setup:**
 
@@ -379,24 +379,41 @@ Add AgentDB to your Claude Desktop config (`~/.config/claude/claude_desktop_conf
 **Available MCP Tools (29 total - v1.3.0):**
 
 *Core Vector DB Tools (5):*
-- `agentdb_init`, `agentdb_insert`, `agentdb_insert_batch`, `agentdb_search`, `agentdb_delete`
+- `agentdb_init` - Initialize database with schema
+- `agentdb_insert` - Insert single vector with metadata
+- `agentdb_insert_batch` - Batch insert with transactions (141x faster)
+- `agentdb_search` - Semantic k-NN vector search with filters
+- `agentdb_delete` - Delete vectors by ID or filters
 
 *Core AgentDB Tools (5 - NEW v1.3.0):*
-- `agentdb_stats`, `agentdb_pattern_store`, `agentdb_pattern_search`, `agentdb_pattern_stats`, `agentdb_clear_cache`
+- `agentdb_stats` - Comprehensive database statistics
+- `agentdb_pattern_store` - Store reasoning patterns with embeddings
+- `agentdb_pattern_search` - Search reasoning patterns semantically
+- `agentdb_pattern_stats` - Pattern analytics and top task types
+- `agentdb_clear_cache` - Cache management for optimal performance
 
 *Frontier Memory Tools (9):*
-- `reflexion_store`, `reflexion_retrieve` - Learn from experience
-- `skill_create`, `skill_search` - Reusable code patterns
-- `causal_add_edge`, `causal_query` - Cause-effect relationships
+- `reflexion_store` - Store episode with self-critique
+- `reflexion_retrieve` - Retrieve relevant past episodes
+- `skill_create` - Create reusable skill
+- `skill_search` - Search for applicable skills
+- `causal_add_edge` - Add causal relationship
+- `causal_query` - Query causal effects
 - `recall_with_certificate` - Utility-based retrieval with provenance
-- `learner_discover` - Automated pattern discovery
-- `db_stats` - Database statistics
+- `learner_discover` - Automated causal pattern discovery
+- `db_stats` - Database statistics showing record counts
 
 *Learning System Tools (10 - NEW v1.3.0):*
-- `learning_start_session`, `learning_end_session` - Session management
-- `learning_predict`, `learning_feedback`, `learning_train` - Adaptive learning
-- `learning_metrics`, `learning_transfer`, `learning_explain` - Analytics & transfer
-- `experience_record`, `reward_signal` - Experience replay & reward shaping
+- `learning_start_session` - Start RL session with algorithm selection
+- `learning_end_session` - End session and save learned policy
+- `learning_predict` - Get AI action recommendations
+- `learning_feedback` - Submit action feedback for learning
+- `learning_train` - Train policy with batch learning
+- `learning_metrics` - Get performance metrics and trends
+- `learning_transfer` - Transfer knowledge between tasks
+- `learning_explain` - Explainable AI recommendations
+- `experience_record` - Record tool execution experience
+- `reward_signal` - Calculate reward signals for learning
 
 [📚 Full MCP Tools Guide](docs/MCP_TOOLS.md) | [🔄 Migration Guide v1.3.0](MIGRATION_v1.3.0.md)
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentdb",
-  "version": "1.3.0",
+  "version": "1.3.1",
   "description": "AgentDB - Frontier Memory Features with MCP Integration: Causal reasoning, reflexion memory, skill library, and automated learning. 150x faster vector search. Full Claude Desktop support via Model Context Protocol.",
   "type": "module",
   "main": "dist/index.js",