Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
SENTINEL_JWT_SECRETNoJWT signing key. Must be set to a secure value (≥16 chars, no placeholder words) in production.dev value
SENTINEL_OIDC_ISSUERNoOIDC issuer URL for SSO.
SENTINEL_DATABASE_URLNoDatabase URL. Use postgresql+asyncpg://… in production.sqlite+aiosqlite:///./sentrik.db
SENTINEL_OIDC_ENABLEDNoEnable SSO via OpenID Connect.false
SENTINEL_SANDBOX_MODENoPer-run egress sandbox mode: none, process, or container.none
SENTINEL_OIDC_AUDIENCENoOIDC audience for SSO.
SENTINEL_OTLP_ENDPOINTNoOTLP collector endpoint for shipping traces.
SENTINEL_USE_LANGGRAPHNoDrive the lifecycle through LangGraph with durable SQLite checkpointer.false
SENTINEL_CRAWL_MAX_DEPTHNoActive-crawl maximum depth.3
SENTINEL_CRAWL_MAX_PAGESNoActive-crawl maximum number of pages.25
SENTINEL_LOCAL_LLM_MODELNoModel name to use with the local OpenAI-compatible endpoint.
SENTINEL_STORAGE_BACKENDNoEvidence/report object store backend: db, local, or s3.db
SENTINEL_ANTHROPIC_API_KEYNoEnable live LLM brains (Anthropic). If not set, uses a deterministic rule-based brain.
SENTINEL_LOCAL_LLM_BASE_URLNoBase URL for an OpenAI-compatible local LLM endpoint.
SENTINEL_SECRET_ENCRYPTION_KEYNoEncrypts test-account secrets at rest. Must be set to a secure value in production.ephemeral
SENTINEL_ALLOW_PRIVATE_NETWORKSNoAllow RFC1918/loopback targets. Set to false in production.true
SENTINEL_STEP_APPROVAL_REQUIREDNoHold state-changing/invasive plan steps as awaiting_approval until an operator approves.true
SENTINEL_USE_DEEPAGENTS_PLANNERNoUse Deep Agents re-ranker over the authorized plan.false
SENTINEL_LANGGRAPH_CHECKPOINT_DBNoSQLite database file for LangGraph checkpoints../sentrik_checkpoints.db
SENTINEL_STEP_APPROVAL_WAIT_SECONDSNoHow long a run waits for approvals before proceeding without the held steps.0
SENTINEL_MAX_REQUESTS_PER_ASSESSMENTNoGlobal request ceiling per assessment.5000
SENTINEL_MAX_LLM_TOKENS_PER_ASSESSMENTNoPer-run LLM token spend ceiling. 0 = unbounded.0
SENTINEL_MAX_LLM_COST_USD_PER_ASSESSMENTNoPer-run LLM cost ceiling in USD. 0 = unbounded.0

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

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