Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
HOSTNoBind address (default 0.0.0.0 in production, localhost otherwise)
PORTNoHTTP port (default 3000)3000
AUTH_2FANo1 forces email sign-in codes on, 0 off. Default: on when RESEND_API_KEY is set. Also gates password-change confirmation and self-service reset
BASE_URLNoPublic origin, used for OAuth redirects, invite links and snippets. Required in production.
MAIL_FROMNoSender address (default STMA <noreply@stma.ai>)STMA <noreply@stma.ai>
SITE_MODENoteaser makes the signed-out site pre-launch: the landing page says the platform is an invite-only private beta and points at the MCP docs, and the guide drops the sections about a console a visitor cannot reach. Signed-in members get the full app and the full guide.
AUTH_LOCALNoLocal username+password accounts (default on; 0 disables)
PGLITE_DIRNoEmbedded database directory (default .data/pglite).data/pglite
DEMO_LOGINSNoCredentials printed on the sign-in page of a throwaway environment: email:password[:label], comma separated, up to 8. Never set on production.
EMBEDDED_DBNo1 allows production on the embedded database — single instance, persist packages/server/.data
STMA_HOSTEDNo1 makes plan limits apply. Unset means this is your instance and nothing is metered.
ADMIN_EMAILSNoComma-separated operator emails for /admin. Works alongside ADMIN_USERNAMES.
DATABASE_URLNoPostgres connection string. Unset → embedded PGlite (dev, or prod with EMBEDDED_DB=1). Required in production unless EMBEDDED_DB=1.
SIGNUPS_OPENNo0 closes new local account registration
AUTH_DEV_MODENo1 forces the dev login form. Auto-enabled outside production when OAuth is not configured.
RESEND_API_KEYNoResend API key for account emails (sign-in codes, password reset). Without it codes are only logged and email 2FA defaults off
ADMIN_USERNAMESNoComma-separated operator usernames for /admin. Unset → /admin is a plain 404.
GITHUB_CLIENT_IDNoOptional GitHub OAuth client ID; callback URL is {BASE_URL}/auth/github/callback
SESSION_TTL_DAYSNoWeb session lifetime (default 30)30
AGENT_STALE_MINUTESNoMark active agent runs stale after no heartbeat (default 3)3
NOTIFY_MAX_PER_HOURNoHard cap on notification emails per person per hour (default 6)6
ERROR_RETENTION_DAYSNoPurge operator error-log entries older than this (default 30; 0 disables the age purge)30
GITHUB_CLIENT_SECRETNoOptional GitHub OAuth client secret
SESSION_RETENTION_DAYSNoPurge resolved sessions older than this (default 0 = keep the archive forever)0
ACTIVITY_RETENTION_DAYSNoPurge activity events, the agent run trail (agent_events) and announcements older than this (default 180; 0 disables the age purge). Ignored for the first two when STMA_HOSTED=1.180
NOTIFY_DEBOUNCE_SECONDSNoWait this long before emailing about a thread so a burst of replies becomes one message (default 120)120
SNAPSHOT_RETENTION_DAYSNoPurge snapshots older than this (default 90, 0 disables). Also bounds stored preflight results.90
AGENT_CLAIM_LEASE_MINUTESNoWork-claim lease refreshed by heartbeat (default 5)5

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

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/Matte-AI-LLC/stma'

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