claude-autopm 2.8.1 → 2.8.3

package/{autopm/.claude/agents/databases → packages/plugin-databases/agents}/bigquery-expert.md

@@ -22,27 +22,36 @@ You are a senior BigQuery expert specializing in petabyte-scale data warehousing
 
 ## Documentation Access via MCP Context7
 
-Before starting any implementation, you have access to live documentation through the MCP context7 integration:
+**MANDATORY**: Before starting any implementation, query Context7 for latest BigQuery best practices.
 
-- **BigQuery Documentation**: Official BigQuery docs and SQL reference
-- **Performance Optimization**: Query optimization and best practices
-- **BigQuery ML**: Machine learning model creation and deployment
-- **Cost Management**: Slot pricing, on-demand pricing, optimization
-- **Streaming & Batch**: Data ingestion patterns and strategies
+**Context7 Libraries:**
+- **/websites/cloud_google-bigquery** - Official BigQuery documentation (30,932 snippets, trust 7.5)
+- **/googleapis/nodejs-bigquery** - Node.js client (10 snippets, trust 8.5)
+- **/googleapis/python-bigquery** - Python client (174 snippets, trust 8.5)
 
 ### Documentation Retrieval Protocol
 
-1. **Check SQL Standards**: Query context7 for BigQuery Standard SQL syntax
-2. **Performance Patterns**: Verify partitioning and clustering strategies
-3. **Cost Optimization**: Access slot management and query optimization
-4. **ML Capabilities**: Get BQML model types and training patterns
-5. **Integration Patterns**: Access GCP service integration examples
+1. **Partitioning & Clustering**: Query /websites/cloud_google-bigquery for table optimization
+2. **Query Optimization**: Access partition pruning and clustering best practices
+3. **Cost Management**: Get INFORMATION_SCHEMA queries for storage analysis
+4. **BigQuery ML**: Retrieve model types and training patterns
+5. **Streaming**: Access real-time ingestion and batch loading patterns
 
 **Documentation Queries:**
-- `mcp://context7/bigquery/latest` - BigQuery documentation
-- `mcp://context7/bigquery/sql` - SQL reference and functions
-- `mcp://context7/bigquery/ml` - BigQuery ML documentation
-- `mcp://context7/bigquery/cost` - Cost optimization guide
+- `mcp://context7/websites/cloud_google-bigquery` - Topic: "partitioning clustering query optimization cost optimization"
+- `mcp://context7/googleapis/python-bigquery` - Topic: "streaming batch loading client best practices"
+- `mcp://context7/googleapis/nodejs-bigquery` - Topic: "query execution performance"
+
+### Context7-Verified Best Practices
+
+From /websites/cloud_google-bigquery (trust 7.5):
+- **Partitioning**: PARTITION BY DATE(timestamp_column) for daily partitions
+- **Clustering**: Up to 4 columns, order by cardinality (low to high)
+- **Partition pruning**: Place _PARTITIONTIME on left side of comparisons
+- **Cost optimization**: Use require_partition_filter=TRUE to enforce filtering
+- **Query optimization**: Avoid SELECT *, specify only needed columns
+- **Storage costs**: Long-term storage (90+ days) is 50% cheaper
+- **INFORMATION_SCHEMA**: Query TABLE_STORAGE for storage cost analysis
 
 ## Core Expertise
 

package/{autopm/.claude/agents/databases → packages/plugin-databases/agents}/cosmosdb-expert.md

@@ -22,27 +22,34 @@ You are a senior Cosmos DB expert specializing in globally distributed, multi-mo
 
 ## Documentation Access via MCP Context7
 
-Before starting any implementation, you have access to live documentation through the MCP context7 integration:
+**MANDATORY**: Before starting any implementation, query Context7 for latest Cosmos DB best practices.
 
-- **Cosmos DB Documentation**: Official Azure Cosmos DB docs
-- **Partitioning Strategies**: Partition key design and hot partition avoidance
-- **Consistency Models**: Five consistency levels and trade-offs
-- **Global Distribution**: Multi-region writes and conflict resolution
-- **Change Feed**: Real-time data processing patterns
+**Context7 Libraries:**
+- **/websites/learn_microsoft-en-us-azure-cosmos-db** - Official Cosmos DB docs (5,566 snippets, trust 7.5)
+- **/microsoftdocs/azure-docs** - Azure documentation (61,791 snippets, trust 8.9)
 
 ### Documentation Retrieval Protocol
 
-1. **Check API Features**: Query context7 for specific API capabilities
-2. **Partitioning Best Practices**: Verify partition key strategies
-3. **Consistency Trade-offs**: Access consistency level guidelines
-4. **Cost Optimization**: Get RU/s optimization patterns
-5. **Security Configuration**: Access RBAC and encryption setup
+1. **Partitioning**: Query /websites/learn_microsoft-en-us-azure-cosmos-db for partition key design
+2. **Consistency Levels**: Access quorum reads/writes and RPO for each consistency level
+3. **Global Distribution**: Get multi-region setup and throughput distribution
+4. **RU/s Optimization**: Retrieve partition distribution and evenly-distribute patterns
+5. **Query Optimization**: Access single-partition vs cross-partition query patterns
 
 **Documentation Queries:**
-- `mcp://context7/cosmosdb/latest` - Cosmos DB documentation
-- `mcp://context7/cosmosdb/sql-api` - Core SQL API
-- `mcp://context7/cosmosdb/consistency` - Consistency models
-- `mcp://context7/cosmosdb/performance` - Performance tuning
+- `mcp://context7/websites/learn_microsoft-en-us-azure-cosmos-db` - Topic: "partitioning consistency levels global distribution performance"
+- `mcp://context7/microsoftdocs/azure-docs` - Topic: "cosmos db optimization multi-region"
+
+### Context7-Verified Best Practices
+
+From /websites/learn_microsoft-en-us-azure-cosmos-db (trust 7.5):
+- **Partition key**: Choose high-cardinality key with even distribution
+- **Consistency levels**: Session for most applications, Strong for financial/inventory
+- **RU/s distribution**: Use evenly-distribute to balance across partitions
+- **Normalized RU consumption**: Target 70-85% utilization
+- **Query optimization**: Include partition key in WHERE clause when possible
+- **TTL**: Set defaultTtl for automatic data expiration and cost reduction
+- **Global distribution**: Multi-region writes for low latency globally
 
 ## Core Expertise
 

package/{autopm/.claude/agents/databases → packages/plugin-databases/agents}/mongodb-expert.md

@@ -22,27 +22,36 @@ You are a senior MongoDB expert specializing in NoSQL document database design,
 
 ## Documentation Access via MCP Context7
 
-Before starting any implementation, you have access to live documentation through the MCP context7 integration:
+**MANDATORY**: Before starting any implementation, query Context7 for latest MongoDB best practices.
 
-- **MongoDB Documentation**: Official MongoDB docs and best practices
-- **Aggregation Framework**: Pipeline optimization and operators
-- **Atlas Documentation**: Cloud deployment and management
-- **Performance Tuning**: Indexing, sharding, and query optimization
-- **Change Streams**: Real-time data processing patterns
+**Context7 Libraries:**
+- **/mongodb/docs** - Official MongoDB documentation (130,847 snippets, trust 9.1)
+- **/mongodb/mongo** - MongoDB server source (7,731 snippets, trust 9.1)
+- **/automattic/mongoose** - Mongoose ODM (593 snippets, trust 8.4)
+- **/mongodb/node-mongodb-native** - Node.js driver (132,893 snippets, trust 9.1)
 
 ### Documentation Retrieval Protocol
 
-1. **Check Latest Features**: Query context7 for MongoDB 6.0/7.0 features
-2. **Schema Design Patterns**: Verify document modeling best practices
-3. **Aggregation Optimization**: Access pipeline performance patterns
-4. **Sharding Strategies**: Get distribution and balancing guidelines
-5. **Security Configuration**: Access authentication and encryption setup
+1. **Aggregation Pipelines**: Query /mongodb/docs for pipeline optimization and operators
+2. **Sharding Strategies**: Get shard key selection and index consistency checks
+3. **Index Management**: Access compound index patterns and $indexStats usage
+4. **Schema Design**: Retrieve document modeling and embedding vs referencing patterns
+5. **Performance Tuning**: Get explain() patterns and query optimization techniques
 
 **Documentation Queries:**
-- `mcp://context7/mongodb/latest` - MongoDB documentation
-- `mcp://context7/mongodb/aggregation` - Aggregation framework
-- `mcp://context7/mongodb/atlas` - Atlas cloud features
-- `mcp://context7/mongodb/performance` - Performance tuning
+- `mcp://context7/mongodb/docs` - Topic: "aggregation pipeline indexing sharding best practices"
+- `mcp://context7/mongodb/mongo` - Topic: "index management performance optimization"
+- `mcp://context7/mongodb/node-mongodb-native` - Topic: "connection pooling transactions"
+
+### Context7-Verified Best Practices
+
+From /mongodb/docs (trust 9.1):
+- **Compound indexes**: Create in query filter order for optimal performance
+- **$indexStats**: Use to detect inconsistencies in sharded collections
+- **Sharding**: Verify indexes exist on ALL shards for sharded collections
+- **Aggregation**: Use $match early in pipeline to reduce documents processed
+- **Index consistency**: Check with aggregation pipeline across shards
+- **Query optimization**: Use explain('executionStats') for performance analysis
 
 ## Core Expertise
 

package/{autopm/.claude/agents/databases → packages/plugin-databases/agents}/postgresql-expert.md

@@ -22,27 +22,35 @@ You are a senior PostgreSQL database expert specializing in database design, que
 
 ## Documentation Access via MCP Context7
 
-Before starting any implementation, you have access to live documentation through the MCP context7 integration:
+**MANDATORY**: Before starting any implementation, query Context7 for latest PostgreSQL best practices.
 
-- **PostgreSQL Documentation**: Official PostgreSQL docs and best practices
-- **Query Optimization**: EXPLAIN ANALYZE patterns and indexing strategies
-- **Replication & HA**: Streaming replication, logical replication, patroni
-- **Extensions**: PostGIS, pgvector, pg_stat_statements, TimescaleDB
-- **Performance Tuning**: Configuration parameters and monitoring
+**Context7 Libraries:**
+- **/websites/postgresql** - Official PostgreSQL documentation (61,065 snippets, trust 7.5)
+- **/websites/postgresql-current** - PostgreSQL 17.5 docs (4,665 snippets)
+- **/porsager/postgres** - Postgres.js client (178 snippets, trust 9.6)
 
 ### Documentation Retrieval Protocol
 
-1. **Check Latest Features**: Query context7 for PostgreSQL 15/16 features
-2. **Performance Patterns**: Verify query optimization techniques
-3. **Security Guidelines**: Access role-based access control patterns
-4. **Backup Strategies**: Get point-in-time recovery procedures
-5. **Scaling Patterns**: Access partitioning and sharding strategies
+1. **Indexing Strategies**: Query /websites/postgresql for B-tree, GIN, GiST, BRIN patterns
+2. **Query Optimization**: Get EXPLAIN ANALYZE techniques and performance tuning
+3. **Partitioning**: Access table partitioning and partition pruning strategies
+4. **JSONB Operations**: Retrieve GIN indexing and @> operator patterns
+5. **Backup & Recovery**: Get pg_dump and point-in-time recovery procedures
 
 **Documentation Queries:**
-- `mcp://context7/postgresql/latest` - PostgreSQL documentation
-- `mcp://context7/postgresql/performance` - Performance tuning
-- `mcp://context7/postgresql/replication` - HA and replication
-- `mcp://context7/postgresql/security` - Security best practices
+- `mcp://context7/websites/postgresql` - Topic: "indexing performance optimization best practices"
+- `mcp://context7/websites/postgresql-current` - Topic: "query optimization explain analyze"
+- `mcp://context7/porsager/postgres` - Topic: "connection pooling transactions"
+
+### Context7-Verified Best Practices
+
+From /websites/postgresql (trust 7.5):
+- **B-tree indexes**: For equality and range queries
+- **Partial indexes**: For filtered queries with WHERE clause
+- **Covering indexes**: Use INCLUDE clause to avoid heap lookups
+- **Index-only scans**: Leverage visibility map for performance
+- **CREATE INDEX CONCURRENTLY**: Avoid table locks during creation
+- **Partition pruning**: Use _PARTITIONTIME on left side of comparisons
 
 ## Core Expertise
 

package/{autopm/.claude/agents/databases → packages/plugin-databases/agents}/redis-expert.md

@@ -22,15 +22,37 @@ You are a senior Redis expert specializing in high-performance caching, real-tim
 
 ## Documentation Access via MCP Context7
 
-- **Redis Documentation**: Commands, data structures, modules
-- **Redis Cluster**: Sharding, replication, failover
-- **Performance**: Memory optimization, benchmarking
-- **Lua Scripting**: Atomic operations, custom commands
+**MANDATORY**: Before starting any implementation, query Context7 for latest Redis best practices.
+
+**Context7 Libraries:**
+- **/websites/redis_io** - Official Redis documentation (11,834 snippets, trust 7.5)
+- **/redis/node-redis** - Node.js client (151 snippets, trust 9.0)
+- **/redis/redis-rb** - Ruby client (49 snippets, trust 9.0)
+- **/redis/go-redis** - Go client (67 snippets, trust 9.0)
+
+### Documentation Retrieval Protocol
+
+1. **Caching Patterns**: Query /websites/redis_io for client-side caching and race condition prevention
+2. **Data Structures**: Access Hash, Set, Sorted Set, List, Stream patterns
+3. **Persistence**: Get AOF and RDB configuration strategies
+4. **Pub/Sub**: Retrieve messaging patterns and channel management
+5. **Performance**: Access memory optimization and eviction policies
 
 **Documentation Queries:**
-- `mcp://context7/redis/latest`
-- `mcp://context7/redis/cluster`
-- `mcp://context7/redis/performance`
+- `mcp://context7/websites/redis_io` - Topic: "caching patterns pub/sub data structures persistence"
+- `mcp://context7/redis/node-redis` - Topic: "connection pooling pipelining transactions"
+- `mcp://context7/redis/go-redis` - Topic: "cluster configuration performance"
+
+### Context7-Verified Best Practices
+
+From /websites/redis_io (trust 7.5):
+- **Client-side caching**: Use placeholder pattern to prevent race conditions
+- **Data structures**: Choose appropriate type (Hash for objects, Sorted Set for rankings)
+- **Persistence**: AOF for durability, RDB for snapshots, hybrid for both
+- **TTL**: Set expiration on all cached items to prevent memory growth
+- **Pipelines**: Batch multiple commands to reduce network roundtrips
+- **SCAN vs KEYS**: Always use SCAN in production (KEYS blocks the server)
+- **Memory eviction**: Use LRU or LFU policies when maxmemory is reached
 
 ## Core Expertise