Skip to main content
Glama
Rinkia

bastiongate

by Rinkia

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
HF_HUB_OFFLINENoSet to 1 for air-gapped hosts to prevent fetching models from HuggingFace; the model must be pre-cached locally.
ANTHROPIC_API_KEYNoAPI key for the Anthropic LLM judge used with agentbastion's inspector_judge option.
BASTIONGATE_EMBED_URLNoSelf-hosted embeddings endpoint URL for semantic detection. Result text is POSTed to this endpoint. Used when BASTIONGATE_EMBED_MODEL is not set.
BASTIONGATE_PROXY_KEYNoOptional key for HTTP proxy mode. When set (or --auth-key is passed), requires an X-Bastiongate-Key header on every request. The key is compared in constant time and never forwarded upstream.
BASTIONGATE_EMBED_WARMNoSet to 0 to defer loading the embedding model until first result instead of at startup. Default is 1 (true).1
BASTIONGATE_EMBED_MODELNoLocal sentence-transformers model for semantic detection (e.g. all-MiniLM-L6-v2). Result text never leaves the process. Preferred over BASTIONGATE_EMBED_URL.
BASTIONGATE_JUDGE_TIMEOUTNoTimeout in seconds for the Anthropic LLM judge. Default is 10s.10
BASTIONGATE_EMBED_INIT_TIMEOUTNoTimeout in seconds for loading the embedding model and templates at gate startup. Default is 120s.120

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