io.github.alex-feel

mcp-context-server

An MCP server that provides persistent multimodal context storage for LLM agents.

communityinfra

Package Details

mcp-context-server

2.2.2•https://pypi.org/project/mcp-context-server
Transportstdio

Environment Variables

LOG_LEVEL(str)

Log level

STORAGE_BACKEND(str)

Storage backend type: sqlite (default) or postgresql

MAX_IMAGE_SIZE_MB(num)

Maximum individual image size in megabytes

MAX_TOTAL_SIZE_MB(num)

Maximum total request size in megabytes

DB_PATH(str)

Custom database file location path

POOL_MAX_READERS(num)

Maximum number of concurrent read connections in the pool

POOL_MAX_WRITERS(num)

Maximum number of concurrent write connections in the pool

POOL_CONNECTION_TIMEOUT_S(num)

Connection timeout in seconds

POOL_IDLE_TIMEOUT_S(num)

Idle connection timeout in seconds

POOL_HEALTH_CHECK_INTERVAL_S(num)

Connection health check interval in seconds

RETRY_MAX_RETRIES(num)

Maximum number of retry attempts for failed operations

RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts

RETRY_MAX_DELAY_S(num)

Maximum delay in seconds between retry attempts

RETRY_JITTER(bool)

Enable random jitter in retry delays

RETRY_BACKOFF_FACTOR(num)

Exponential backoff multiplication factor for retries

SQLITE_FOREIGN_KEYS(bool)

Enable SQLite foreign key constraints

SQLITE_JOURNAL_MODE(str)

SQLite journal mode (e.g., WAL, DELETE)

SQLITE_SYNCHRONOUS(str)

SQLite synchronous mode (e.g., NORMAL, FULL, OFF)

SQLITE_TEMP_STORE(str)

SQLite temporary storage location (e.g., MEMORY, FILE)

SQLITE_MMAP_SIZE(num)

SQLite memory-mapped I/O size in bytes

SQLITE_CACHE_SIZE(num)

SQLite cache size (negative value for KB, positive for pages)

SQLITE_PAGE_SIZE(num)

SQLite page size in bytes

SQLITE_WAL_AUTOCHECKPOINT(num)

SQLite WAL autocheckpoint threshold in pages

SQLITE_BUSY_TIMEOUT_MS(num)

SQLite busy timeout in milliseconds

SQLITE_WAL_CHECKPOINT(str)

SQLite WAL checkpoint mode (e.g., PASSIVE, FULL, RESTART)

SHUTDOWN_TIMEOUT_S(num)

Server shutdown timeout in seconds

SHUTDOWN_TIMEOUT_TEST_S(num)

Test mode shutdown timeout in seconds

QUEUE_TIMEOUT_S(num)

Queue operation timeout in seconds

QUEUE_TIMEOUT_TEST_S(num)

Test mode queue timeout in seconds

CIRCUIT_BREAKER_FAILURE_THRESHOLD(num)

Circuit breaker failure threshold before opening

CIRCUIT_BREAKER_RECOVERY_TIMEOUT_S(num)

Circuit breaker recovery timeout in seconds

CIRCUIT_BREAKER_HALF_OPEN_MAX_CALLS(num)

Maximum calls allowed in circuit breaker half-open state

POSTGRESQL_CONNECTION_STRING(str)
Secret

Complete PostgreSQL connection string (overrides individual settings if provided)

POSTGRESQL_HOST(str)

PostgreSQL server host address

POSTGRESQL_PORT(num)

PostgreSQL server port number

POSTGRESQL_USER(str)

PostgreSQL database username

POSTGRESQL_PASSWORD(str)
Secret

PostgreSQL database password

POSTGRESQL_DATABASE(str)

PostgreSQL database name

POSTGRESQL_POOL_MIN(num)

PostgreSQL connection pool minimum size

POSTGRESQL_POOL_MAX(num)

PostgreSQL connection pool maximum size

POSTGRESQL_POOL_TIMEOUT_S(num)

PostgreSQL connection pool timeout in seconds

POSTGRESQL_COMMAND_TIMEOUT_S(num)

PostgreSQL command execution timeout in seconds

POSTGRESQL_MIGRATION_TIMEOUT_S(num)

Timeout in seconds for PostgreSQL migration operations (default: 300)

POSTGRESQL_MAX_INACTIVE_LIFETIME_S(num)

Close idle PostgreSQL connections after this many seconds (0 to disable, default: 300)

POSTGRESQL_MAX_QUERIES(num)

Recycle PostgreSQL connections after this many queries (0 to disable, default: 10000)

POSTGRESQL_TCP_KEEPALIVES_IDLE_S(num)

Seconds of idle time before sending first TCP keepalive probe (0 to disable, default: 15)

POSTGRESQL_TCP_KEEPALIVES_INTERVAL_S(num)

Seconds between subsequent TCP keepalive probes (0 to disable, default: 5)

POSTGRESQL_TCP_KEEPALIVES_COUNT(num)

Number of failed TCP keepalive probes before connection is considered dead (0 to disable, default: 3)

POSTGRESQL_STATEMENT_CACHE_SIZE(num)

asyncpg prepared statement cache size. Set to 0 for external pooler compatibility (PgBouncer transaction mode, Pgpool-II, etc.). Default: 100

POSTGRESQL_MAX_CACHED_STATEMENT_LIFETIME_S(num)

Maximum lifetime of cached prepared statements in seconds (default: 300). Has no effect when statement_cache_size=0

POSTGRESQL_MAX_CACHEABLE_STATEMENT_SIZE(num)

Maximum size of statement to cache in bytes (default: 15360). Has no effect when statement_cache_size=0

POSTGRESQL_SSL_MODE(str)

PostgreSQL SSL mode (disable, allow, prefer, require, verify-ca, verify-full)

POSTGRESQL_SCHEMA(str)

PostgreSQL schema name for table and index operations (default: public)

ENABLE_SEMANTIC_SEARCH(bool)

Enable semantic search functionality

ENABLE_EMBEDDING_GENERATION(bool)

Enable embedding generation for stored context. Default true - server fails if dependencies not met. Set false to disable embeddings.

OLLAMA_HOST(str)

Ollama API host URL for embedding generation

OLLAMA_AUTO_PULL(bool)

Automatically pull missing Ollama models on startup (default: true)

OLLAMA_PULL_TIMEOUT_S(num)

Timeout in seconds for pulling Ollama models (default: 900, range: 30-3600)

EMBEDDING_OLLAMA_TRUNCATE(bool)

Ollama embedding truncation mode: false (default) returns error when context exceeded, true enables silent truncation

EMBEDDING_OLLAMA_NUM_CTX(num)

Ollama embedding context window size in tokens (default: 4096, range: 512-2097152)

EMBEDDING_MODEL(str)

Embedding model name for semantic search

EMBEDDING_DIM(num)

Embedding vector dimensions

EMBEDDING_TIMEOUT_S(num)

Timeout in seconds for embedding generation API calls

EMBEDDING_RETRY_MAX_ATTEMPTS(num)

Maximum number of retry attempts for embedding generation

EMBEDDING_RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts (with exponential backoff)

EMBEDDING_MAX_CONCURRENT(num)

Maximum concurrent embedding generation operations (default: 3, range: 1-20)

ENABLE_SUMMARY_GENERATION(bool)

Enable summary generation for stored context. Default true - server fails if dependencies not met. Set false to disable summaries.

SUMMARY_PROVIDER(str)

Summary provider: ollama (default), openai, or anthropic

SUMMARY_MODEL(str)

Summary generation model name (default: qwen3:0.6b)

SUMMARY_MAX_TOKENS(num)

Maximum output tokens for summary generation (default: 4000, range: 50-16384). Increase if summaries are truncated by reasoning models

SUMMARY_TIMEOUT_S(num)

Timeout in seconds for summary generation API calls

SUMMARY_RETRY_MAX_ATTEMPTS(num)

Maximum number of retry attempts for summary generation

SUMMARY_RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts (with exponential backoff)

SUMMARY_MAX_CONCURRENT(num)

Maximum concurrent summary generation operations (default: 3, range: 1-20)

SUMMARY_PROMPT(str)

Custom summarization prompt. Overrides the built-in default. Used as system message for the LLM.

SUMMARY_MIN_CONTENT_LENGTH(num)

Minimum text content length in characters to trigger summary generation (default: 500, range: 0-10000). Set to 0 to always generate.

SUMMARY_OLLAMA_NUM_CTX(num)

Ollama summary context window size in tokens (default: 32768, range: 512-2097152)

SUMMARY_OLLAMA_TRUNCATE(bool)

Ollama summary truncation mode: false (default) returns error when context exceeded, true enables silent truncation

SUMMARY_OPENAI_REASONING_EFFORT(str)

Reasoning effort level for OpenAI reasoning models (default: low). Valid values vary by generation: gpt-5: low, medium, high; gpt-5.1+: none, low, medium, high, xhigh. Default low is universally valid across all generations

SUMMARY_ANTHROPIC_EFFORT(str)

Effort level for Anthropic Claude models (default: none). Valid values: max, high, medium, low. Controls inference effort (adaptive thinking)

ANTHROPIC_API_KEY(str)
Secret

Anthropic API key for summary generation

ENABLE_FTS(bool)

Enable full-text search functionality

FTS_LANGUAGE(str)

Language for FTS stemming (e.g., english, german, french)

FTS_RERANK_WINDOW_SIZE(num)

Characters of context around each FTS match for reranking passage extraction (default: 750)

FTS_RERANK_GAP_MERGE(num)

Merge FTS match regions within this character distance (default: 100)

ENABLE_HYBRID_SEARCH(bool)

Enable hybrid search combining FTS and semantic search with RRF fusion

HYBRID_RRF_K(num)

RRF smoothing constant for hybrid search (default 60)

HYBRID_RRF_OVERFETCH(num)

Multiplier for over-fetching results before RRF fusion (default: 2)

HYBRID_FTS_OR_THRESHOLD(num)

Minimum significant query terms to switch hybrid FTS from AND to OR logic (default: 4)

SEARCH_DEFAULT_SORT_BY(str)

Default sort order for search results: relevance (only 'relevance' supported in current version)

SEARCH_TRUNCATION_LENGTH(num)

Maximum character length for truncated text_content in search results (default: 300, range: 50-1000)

ENABLE_CHUNKING(bool)

Enable text chunking for embedding generation (default: true)

CHUNK_SIZE(num)

Target chunk size in characters (default: 1500)

CHUNK_OVERLAP(num)

Overlap between chunks in characters (default: 150)

CHUNK_AGGREGATION(str)

Chunk score aggregation method: max (only 'max' supported in current version)

CHUNK_DEDUP_OVERFETCH(num)

Multiplier for over-fetching chunks before deduplication (default: 5)

ENABLE_RERANKING(bool)

Enable cross-encoder reranking of search results (default: true)

RERANKING_PROVIDER(str)

Reranking provider (default: flashrank)

RERANKING_MODEL(str)

Reranking model name (default: ms-marco-MiniLM-L-12-v2)

RERANKING_MAX_LENGTH(num)

Maximum input length for reranking in tokens (default: 512)

RERANKING_OVERFETCH(num)

Multiplier for over-fetching results before reranking (default: 4)

RERANKING_CACHE_DIR(str)

Directory for caching reranking models

RERANKING_CHARS_PER_TOKEN(num)

Estimated characters per token for passage size validation (default: 4.0, range: 2.0-8.0)

RERANKING_INTRA_OP_THREADS(num)

ONNX Runtime intra-operation parallelism threads for reranking (default: 0 = auto-detect)

RERANKING_CPU_MEM_ARENA(bool)

Enable ONNX Runtime CPU memory arena for reranking (default: false)

RERANKING_BATCH_SIZE(num)

Maximum passages per ONNX Runtime inference batch during reranking (default: 32)

EMBEDDING_PROVIDER(str)

Embedding provider: ollama (default), openai, azure, huggingface, or voyage

OPENAI_API_KEY(str)
Secret

OpenAI API key for OpenAI embedding provider

OPENAI_API_BASE(str)

Custom base URL for OpenAI-compatible APIs

OPENAI_ORGANIZATION(str)

OpenAI organization ID

AZURE_OPENAI_API_KEY(str)
Secret

Azure OpenAI API key

AZURE_OPENAI_ENDPOINT(str)

Azure OpenAI endpoint URL

AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME(str)

Azure OpenAI embedding deployment name

AZURE_OPENAI_API_VERSION(str)

Azure OpenAI API version (default: 2024-02-01)

HUGGINGFACEHUB_API_TOKEN(str)
Secret

HuggingFace Hub API token for HuggingFace embedding provider

VOYAGE_API_KEY(str)
Secret

Voyage AI API key for Voyage embedding provider

VOYAGE_TRUNCATION(bool)

Voyage AI truncation mode: false (default) returns error when context exceeded, true enables silent truncation

VOYAGE_BATCH_SIZE(num)

Voyage AI batch size for embedding requests

LANGSMITH_TRACING(bool)

Enable LangSmith tracing

LANGSMITH_API_KEY(str)
Secret

LangSmith API key

LANGSMITH_PROJECT(str)

LangSmith project name

LANGSMITH_ENDPOINT(str)

LangSmith API endpoint URL

METADATA_INDEXED_FIELDS(str)

Comma-separated list of metadata fields to index (field:type format)

METADATA_INDEX_SYNC_MODE(str)

Index sync mode: strict (fail), auto (sync), warn (log), additive (default, add missing only)

MCP_TRANSPORT(str)

Transport mode: stdio for local, http for Docker/remote

FASTMCP_HOST(str)

HTTP bind address (use 0.0.0.0 for Docker)

FASTMCP_PORT(num)

HTTP port number

FASTMCP_STATELESS_HTTP(bool)

Enable stateless HTTP mode for horizontal scaling. Enabled by default as the server has no stateful MCP features. Set to false only if you need server-side MCP session tracking.

DISABLED_TOOLS(str)

Comma-separated list of tools to disable (e.g., delete_context,update_context)

MCP_AUTH_TOKEN(str)
Secret

Bearer token for HTTP authentication (required when using SimpleTokenVerifier)

MCP_AUTH_CLIENT_ID(str)

Client ID to assign to authenticated requests

MCP_AUTH_PROVIDER(str)

Authentication provider: none (default), simple_token

MCP_SERVER_INSTRUCTIONS(str)

Custom server instructions text. Overrides built-in default. Set to empty string to disable.

ghcr.io/alex-feel/mcp-context-server:2.2.2

https://github.com/alex-feel/mcp-context-server/pkgs/container/mcp-context-server
TransportStreamable HTTP
Runtimedocker

Environment Variables

LOG_LEVEL(str)

Log level

STORAGE_BACKEND(str)

Storage backend type: sqlite (default) or postgresql

MAX_IMAGE_SIZE_MB(num)

Maximum individual image size in megabytes

MAX_TOTAL_SIZE_MB(num)

Maximum total request size in megabytes

DB_PATH(str)

Custom database file location path

POOL_MAX_READERS(num)

Maximum number of concurrent read connections in the pool

POOL_MAX_WRITERS(num)

Maximum number of concurrent write connections in the pool

POOL_CONNECTION_TIMEOUT_S(num)

Connection timeout in seconds

POOL_IDLE_TIMEOUT_S(num)

Idle connection timeout in seconds

POOL_HEALTH_CHECK_INTERVAL_S(num)

Connection health check interval in seconds

RETRY_MAX_RETRIES(num)

Maximum number of retry attempts for failed operations

RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts

RETRY_MAX_DELAY_S(num)

Maximum delay in seconds between retry attempts

RETRY_JITTER(bool)

Enable random jitter in retry delays

RETRY_BACKOFF_FACTOR(num)

Exponential backoff multiplication factor for retries

SQLITE_FOREIGN_KEYS(bool)

Enable SQLite foreign key constraints

SQLITE_JOURNAL_MODE(str)

SQLite journal mode (e.g., WAL, DELETE)

SQLITE_SYNCHRONOUS(str)

SQLite synchronous mode (e.g., NORMAL, FULL, OFF)

SQLITE_TEMP_STORE(str)

SQLite temporary storage location (e.g., MEMORY, FILE)

SQLITE_MMAP_SIZE(num)

SQLite memory-mapped I/O size in bytes

SQLITE_CACHE_SIZE(num)

SQLite cache size (negative value for KB, positive for pages)

SQLITE_PAGE_SIZE(num)

SQLite page size in bytes

SQLITE_WAL_AUTOCHECKPOINT(num)

SQLite WAL autocheckpoint threshold in pages

SQLITE_BUSY_TIMEOUT_MS(num)

SQLite busy timeout in milliseconds

SQLITE_WAL_CHECKPOINT(str)

SQLite WAL checkpoint mode (e.g., PASSIVE, FULL, RESTART)

SHUTDOWN_TIMEOUT_S(num)

Server shutdown timeout in seconds

SHUTDOWN_TIMEOUT_TEST_S(num)

Test mode shutdown timeout in seconds

QUEUE_TIMEOUT_S(num)

Queue operation timeout in seconds

QUEUE_TIMEOUT_TEST_S(num)

Test mode queue timeout in seconds

CIRCUIT_BREAKER_FAILURE_THRESHOLD(num)

Circuit breaker failure threshold before opening

CIRCUIT_BREAKER_RECOVERY_TIMEOUT_S(num)

Circuit breaker recovery timeout in seconds

CIRCUIT_BREAKER_HALF_OPEN_MAX_CALLS(num)

Maximum calls allowed in circuit breaker half-open state

POSTGRESQL_CONNECTION_STRING(str)
Secret

Complete PostgreSQL connection string (overrides individual settings if provided)

POSTGRESQL_HOST(str)

PostgreSQL server host address

POSTGRESQL_PORT(num)

PostgreSQL server port number

POSTGRESQL_USER(str)

PostgreSQL database username

POSTGRESQL_PASSWORD(str)
Secret

PostgreSQL database password

POSTGRESQL_DATABASE(str)

PostgreSQL database name

POSTGRESQL_POOL_MIN(num)

PostgreSQL connection pool minimum size

POSTGRESQL_POOL_MAX(num)

PostgreSQL connection pool maximum size

POSTGRESQL_POOL_TIMEOUT_S(num)

PostgreSQL connection pool timeout in seconds

POSTGRESQL_COMMAND_TIMEOUT_S(num)

PostgreSQL command execution timeout in seconds

POSTGRESQL_MIGRATION_TIMEOUT_S(num)

Timeout in seconds for PostgreSQL migration operations (default: 300)

POSTGRESQL_MAX_INACTIVE_LIFETIME_S(num)

Close idle PostgreSQL connections after this many seconds (0 to disable, default: 300)

POSTGRESQL_MAX_QUERIES(num)

Recycle PostgreSQL connections after this many queries (0 to disable, default: 10000)

POSTGRESQL_TCP_KEEPALIVES_IDLE_S(num)

Seconds of idle time before sending first TCP keepalive probe (0 to disable, default: 15)

POSTGRESQL_TCP_KEEPALIVES_INTERVAL_S(num)

Seconds between subsequent TCP keepalive probes (0 to disable, default: 5)

POSTGRESQL_TCP_KEEPALIVES_COUNT(num)

Number of failed TCP keepalive probes before connection is considered dead (0 to disable, default: 3)

POSTGRESQL_STATEMENT_CACHE_SIZE(num)

asyncpg prepared statement cache size. Set to 0 for external pooler compatibility (PgBouncer transaction mode, Pgpool-II, etc.). Default: 100

POSTGRESQL_MAX_CACHED_STATEMENT_LIFETIME_S(num)

Maximum lifetime of cached prepared statements in seconds (default: 300). Has no effect when statement_cache_size=0

POSTGRESQL_MAX_CACHEABLE_STATEMENT_SIZE(num)

Maximum size of statement to cache in bytes (default: 15360). Has no effect when statement_cache_size=0

POSTGRESQL_SSL_MODE(str)

PostgreSQL SSL mode (disable, allow, prefer, require, verify-ca, verify-full)

POSTGRESQL_SCHEMA(str)

PostgreSQL schema name for table and index operations (default: public)

ENABLE_SEMANTIC_SEARCH(bool)

Enable semantic search functionality

ENABLE_EMBEDDING_GENERATION(bool)

Enable embedding generation for stored context. Default true - server fails if dependencies not met. Set false to disable embeddings.

OLLAMA_HOST(str)

Ollama API host URL for embedding generation

OLLAMA_AUTO_PULL(bool)

Automatically pull missing Ollama models on startup (default: true)

OLLAMA_PULL_TIMEOUT_S(num)

Timeout in seconds for pulling Ollama models (default: 900, range: 30-3600)

EMBEDDING_OLLAMA_TRUNCATE(bool)

Ollama embedding truncation mode: false (default) returns error when context exceeded, true enables silent truncation

EMBEDDING_OLLAMA_NUM_CTX(num)

Ollama embedding context window size in tokens (default: 4096, range: 512-2097152)

EMBEDDING_MODEL(str)

Embedding model name for semantic search

EMBEDDING_DIM(num)

Embedding vector dimensions

EMBEDDING_TIMEOUT_S(num)

Timeout in seconds for embedding generation API calls

EMBEDDING_RETRY_MAX_ATTEMPTS(num)

Maximum number of retry attempts for embedding generation

EMBEDDING_RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts (with exponential backoff)

EMBEDDING_MAX_CONCURRENT(num)

Maximum concurrent embedding generation operations (default: 3, range: 1-20)

ENABLE_SUMMARY_GENERATION(bool)

Enable summary generation for stored context. Default true - server fails if dependencies not met. Set false to disable summaries.

SUMMARY_PROVIDER(str)

Summary provider: ollama (default), openai, or anthropic

SUMMARY_MODEL(str)

Summary generation model name (default: qwen3:0.6b)

SUMMARY_MAX_TOKENS(num)

Maximum output tokens for summary generation (default: 4000, range: 50-16384). Increase if summaries are truncated by reasoning models

SUMMARY_TIMEOUT_S(num)

Timeout in seconds for summary generation API calls

SUMMARY_RETRY_MAX_ATTEMPTS(num)

Maximum number of retry attempts for summary generation

SUMMARY_RETRY_BASE_DELAY_S(num)

Base delay in seconds between retry attempts (with exponential backoff)

SUMMARY_MAX_CONCURRENT(num)

Maximum concurrent summary generation operations (default: 3, range: 1-20)

SUMMARY_PROMPT(str)

Custom summarization prompt. Overrides the built-in default. Used as system message for the LLM.

SUMMARY_MIN_CONTENT_LENGTH(num)

Minimum text content length in characters to trigger summary generation (default: 500, range: 0-10000). Set to 0 to always generate.

SUMMARY_OLLAMA_NUM_CTX(num)

Ollama summary context window size in tokens (default: 32768, range: 512-2097152)

SUMMARY_OLLAMA_TRUNCATE(bool)

Ollama summary truncation mode: false (default) returns error when context exceeded, true enables silent truncation

SUMMARY_OPENAI_REASONING_EFFORT(str)

Reasoning effort level for OpenAI reasoning models (default: low). Valid values vary by generation: gpt-5: low, medium, high; gpt-5.1+: none, low, medium, high, xhigh. Default low is universally valid across all generations

SUMMARY_ANTHROPIC_EFFORT(str)

Effort level for Anthropic Claude models (default: none). Valid values: max, high, medium, low. Controls inference effort (adaptive thinking)

ANTHROPIC_API_KEY(str)
Secret

Anthropic API key for summary generation

ENABLE_FTS(bool)

Enable full-text search functionality

FTS_LANGUAGE(str)

Language for FTS stemming (e.g., english, german, french)

FTS_RERANK_WINDOW_SIZE(num)

Characters of context around each FTS match for reranking passage extraction (default: 750)

FTS_RERANK_GAP_MERGE(num)

Merge FTS match regions within this character distance (default: 100)

ENABLE_HYBRID_SEARCH(bool)

Enable hybrid search combining FTS and semantic search with RRF fusion

HYBRID_RRF_K(num)

RRF smoothing constant for hybrid search (default 60)

HYBRID_RRF_OVERFETCH(num)

Multiplier for over-fetching results before RRF fusion (default: 2)

HYBRID_FTS_OR_THRESHOLD(num)

Minimum significant query terms to switch hybrid FTS from AND to OR logic (default: 4)

SEARCH_DEFAULT_SORT_BY(str)

Default sort order for search results: relevance (only 'relevance' supported in current version)

SEARCH_TRUNCATION_LENGTH(num)

Maximum character length for truncated text_content in search results (default: 300, range: 50-1000)

ENABLE_CHUNKING(bool)

Enable text chunking for embedding generation (default: true)

CHUNK_SIZE(num)

Target chunk size in characters (default: 1500)

CHUNK_OVERLAP(num)

Overlap between chunks in characters (default: 150)

CHUNK_AGGREGATION(str)

Chunk score aggregation method: max (only 'max' supported in current version)

CHUNK_DEDUP_OVERFETCH(num)

Multiplier for over-fetching chunks before deduplication (default: 5)

ENABLE_RERANKING(bool)

Enable cross-encoder reranking of search results (default: true)

RERANKING_PROVIDER(str)

Reranking provider (default: flashrank)

RERANKING_MODEL(str)

Reranking model name (default: ms-marco-MiniLM-L-12-v2)

RERANKING_MAX_LENGTH(num)

Maximum input length for reranking in tokens (default: 512)

RERANKING_OVERFETCH(num)

Multiplier for over-fetching results before reranking (default: 4)

RERANKING_CACHE_DIR(str)

Directory for caching reranking models

RERANKING_CHARS_PER_TOKEN(num)

Estimated characters per token for passage size validation (default: 4.0, range: 2.0-8.0)

RERANKING_INTRA_OP_THREADS(num)

ONNX Runtime intra-operation parallelism threads for reranking (default: 0 = auto-detect)

RERANKING_CPU_MEM_ARENA(bool)

Enable ONNX Runtime CPU memory arena for reranking (default: false)

RERANKING_BATCH_SIZE(num)

Maximum passages per ONNX Runtime inference batch during reranking (default: 32)

EMBEDDING_PROVIDER(str)

Embedding provider: ollama (default), openai, azure, huggingface, or voyage

OPENAI_API_KEY(str)
Secret

OpenAI API key for OpenAI embedding provider

OPENAI_API_BASE(str)

Custom base URL for OpenAI-compatible APIs

OPENAI_ORGANIZATION(str)

OpenAI organization ID

AZURE_OPENAI_API_KEY(str)
Secret

Azure OpenAI API key

AZURE_OPENAI_ENDPOINT(str)

Azure OpenAI endpoint URL

AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME(str)

Azure OpenAI embedding deployment name

AZURE_OPENAI_API_VERSION(str)

Azure OpenAI API version (default: 2024-02-01)

HUGGINGFACEHUB_API_TOKEN(str)
Secret

HuggingFace Hub API token for HuggingFace embedding provider

VOYAGE_API_KEY(str)
Secret

Voyage AI API key for Voyage embedding provider

VOYAGE_TRUNCATION(bool)

Voyage AI truncation mode: false (default) returns error when context exceeded, true enables silent truncation

VOYAGE_BATCH_SIZE(num)

Voyage AI batch size for embedding requests

LANGSMITH_TRACING(bool)

Enable LangSmith tracing

LANGSMITH_API_KEY(str)
Secret

LangSmith API key

LANGSMITH_PROJECT(str)

LangSmith project name

LANGSMITH_ENDPOINT(str)

LangSmith API endpoint URL

METADATA_INDEXED_FIELDS(str)

Comma-separated list of metadata fields to index (field:type format)

METADATA_INDEX_SYNC_MODE(str)

Index sync mode: strict (fail), auto (sync), warn (log), additive (default, add missing only)

MCP_TRANSPORT(str)

Transport mode: stdio for local, http for Docker/remote

FASTMCP_HOST(str)

HTTP bind address (use 0.0.0.0 for Docker)

FASTMCP_PORT(num)

HTTP port number

FASTMCP_STATELESS_HTTP(bool)

Enable stateless HTTP mode for horizontal scaling. Enabled by default as the server has no stateful MCP features. Set to false only if you need server-side MCP session tracking.

DISABLED_TOOLS(str)

Comma-separated list of tools to disable (e.g., delete_context,update_context)

MCP_AUTH_TOKEN(str)
Secret

Bearer token for HTTP authentication (required when using SimpleTokenVerifier)

MCP_AUTH_CLIENT_ID(str)

Client ID to assign to authenticated requests

MCP_AUTH_PROVIDER(str)

Authentication provider: none (default), simple_token

MCP_SERVER_INSTRUCTIONS(str)

Custom server instructions text. Overrides built-in default. Set to empty string to disable.