Skip to main content
Glama
ABHIJEET-MUNESHWAR

WisdomAgent

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
WA_HOSTNoHost interface to bind the API server to.0.0.0.0
WA_PORTNoPort for the API server.8000
WA_RRF_KNoReciprocal rank fusion constant k.60
WA_BM25_BNoBM25 b parameter.0.75
WA_BM25_K1NoBM25 K1 parameter.1.2
WA_LOG_JSONNoEmit structured JSON logs if 'true', otherwise pretty-print.true
WA_LLM_MODELNoModel name for LLM calls.gpt-4o-mini
WA_LOG_LEVELNoLogging level (DEBUG, INFO, WARNING, ERROR).INFO
WA_MAX_TOP_KNoMaximum allowed top-K (requests are clamped).100
WA_REDIS_URLNoRedis connection URL. Blank disables caching.
WA_MAX_CLAIMSNoMaximum number of claims in an answer.12
WA_NGRAM_SIZENoCharacter n-gram size for tokenization.4
WA_ENVIRONMENTNoEnvironment label for logs and metrics.local
WA_KAFKA_TOPICNoKafka topic for domain events.wisdomagent.events
WA_LLM_API_KEYNoAPI key for the LLM provider.
WA_MIN_SUPPORTNoMinimum support threshold for grounding.0.6
WA_DATABASE_DSNNoPostgres connection string. Blank uses in-memory repository.
WA_LLM_BASE_URLNoBase URL for OpenAI-compatible LLM API. Blank uses deterministic offline path.
WA_DEFAULT_TOP_KNoDefault top-K for retrieval.10
WA_EMBEDDING_DIMNoEmbedding dimension for hashed vectors.128
WA_KAFKA_BROKERSNoComma-separated Kafka broker list. Blank uses in-process publisher.
WA_OTLP_ENDPOINTNoOTLP endpoint for tracing. Blank disables tracing.
WA_TENANT_SHARDSNoNumber of hash shards for tenant partitioning.4
WA_BOOTSTRAP_DEMONoIf set, seeds the corpus with synthetic demo data.
WA_DEFAULT_TENANTNoTenant used when x-tenant-id header is absent.default
WA_EMBEDDING_SEEDNoSeed for deterministic hashed embeddings.1337
WA_EVIDENCE_LIMITNoMaximum number of evidence documents considered.8
WA_AGENT_MAX_STEPSNoMaximum number of agent steps.8
WA_FUSION_STRATEGYNoFusion strategy: 'rrf' or 'weighted'.rrf
WA_TIMEOUT_SECONDSNoTimeout for outbound I/O calls.5.0
WA_AGENT_MAX_TOKENSNoMaximum token budget for agent.8000
WA_MIN_GROUNDEDNESSNoMinimum groundedness score for acceptance.0.7
WA_AGENT_MAX_REPAIRSNoMaximum number of repair attempts.1
WA_CACHE_TTL_SECONDSNoTTL for cached answers/aggregates.300
WA_RETRY_MAX_ATTEMPTSNoMaximum retry attempts for resilient calls.3
WA_RATE_LIMIT_CAPACITYNoRate limiter capacity (burst).200
WA_WORKER_METRICS_PORTNoPort for the worker's metrics endpoint.8001
WA_BREAKER_OPEN_SECONDSNoHow long the circuit breaker stays open.10.0
WA_CANDIDATE_MULTIPLIERNoRetrieval depth multiplier before reranking.4
WA_MCP_MAX_OUTPUT_CHARSNoMaximum characters in MCP tool output.20000
WA_PARTITION_SPAN_MONTHSNoWidth of range partitions in months.1
WA_MIN_EVIDENCE_RELEVANCENoEvidence relevance floor below which the system refuses.0.25
WA_BULKHEAD_MAX_CONCURRENTNoMaximum concurrent calls per dependency.32
WA_AGENT_MAX_CALLS_PER_TOOLNoMaximum number of calls per tool per run.3
WA_MCP_CALL_TIMEOUT_SECONDSNoTimeout for MCP tool calls.10.0
WA_MONITOR_INTERVAL_SECONDSNoMonitor worker sweep interval.300.0
WA_BREAKER_FAILURE_THRESHOLDNoNumber of failures to open the circuit breaker.5
WA_AGENT_MAX_WALL_CLOCK_SECONDSNoMaximum wall clock time for a run.30.0
WA_RATE_LIMIT_REFILL_PER_SECONDNoRate limiter refill rate per second.100.0

Capabilities

Server capabilities have not been inspected yet.

Tools

Functions exposed to the LLM to take actions

NameDescription

No tools

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ABHIJEET-MUNESHWAR/WisdomAgent'

If you have feedback or need assistance with the MCP directory API, please join our Discord server