Skip to main content
Glama
.env.example4.63 kB
# ============================================================================= # Relace MCP Server - Environment Variables # ============================================================================= # Copy this file to .env for local development. # For MCP usage, set these in your mcp_config.json "env" section instead. # ============================================================================= # ----------------------------------------------------------------------------- # Required # ----------------------------------------------------------------------------- # Relace API Key (get yours at: https://app.relace.ai/settings/billing) RELACE_API_KEY=rlc-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # ----------------------------------------------------------------------------- # Recommended # ----------------------------------------------------------------------------- # Restrict file access to this directory (prevents path traversal attacks) # Defaults to current working directory if not set RELACE_BASE_DIR=/path/to/your/project # Strict mode: set to 1 to require explicit RELACE_BASE_DIR (recommended for production) # RELACE_STRICT_MODE=1 # ----------------------------------------------------------------------------- # Advanced: Apply API Settings # ----------------------------------------------------------------------------- # Apply provider (default: relace) # Set to openai/openrouter/cerebras (or any non-"relace" label) to use an OpenAI-compatible chat.completions endpoint. # RELACE_APPLY_PROVIDER=relace # API key selection (in order): # 1) RELACE_APPLY_API_KEY # 2) RELACE_APPLY_API_KEY_ENV -> another env var name (e.g. OPENROUTER_API_KEY) # 3) Provider default env var (e.g. OPENAI_API_KEY / OPENROUTER_API_KEY / CEREBRAS_API_KEY) # # RELACE_APPLY_API_KEY= # RELACE_APPLY_API_KEY_ENV= # # OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # OPENROUTER_API_KEY=sk-or-v1-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # CEREBRAS_API_KEY=csk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # Apply API endpoint (default: Relace official endpoint) # RELACE_APPLY_ENDPOINT=https://instantapply.endpoint.relace.run/v1/apply # Apply model name (Relace default: auto) # RELACE_APPLY_MODEL=auto # Optional JSON object of default headers (passed as OpenAI SDK default_headers) # Example: # RELACE_APPLY_HEADERS={"HTTP-Referer":"https://example.com","X-Title":"relace-mcp"} # Apply request timeout in seconds # RELACE_TIMEOUT_SECONDS=60 # Maximum retry attempts for transient errors # RELACE_MAX_RETRIES=3 # Base delay between retries (exponential backoff) # RELACE_RETRY_BASE_DELAY=1.0 # ----------------------------------------------------------------------------- # Advanced: Search API Settings # ----------------------------------------------------------------------------- # Search provider (default: relace) # Set to openai/openrouter/cerebras (or any non-"relace" label) to use an OpenAI-compatible chat.completions endpoint. # RELACE_SEARCH_PROVIDER=relace # API key selection (in order): # 1) RELACE_SEARCH_API_KEY # 2) RELACE_SEARCH_API_KEY_ENV -> another env var name (e.g. OPENROUTER_API_KEY) # 3) Provider default env var (e.g. OPENAI_API_KEY / OPENROUTER_API_KEY / CEREBRAS_API_KEY) # # RELACE_SEARCH_API_KEY= # RELACE_SEARCH_API_KEY_ENV= # # OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # OPENROUTER_API_KEY=sk-or-v1-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # CEREBRAS_API_KEY=csk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # Search API endpoint (default: Relace official endpoint) # RELACE_SEARCH_ENDPOINT=https://search.endpoint.relace.run/v1/search # Search model name # RELACE_SEARCH_MODEL=relace-search # Search request timeout in seconds # RELACE_SEARCH_TIMEOUT_SECONDS=120 # Maximum agent turns for search # RELACE_SEARCH_MAX_TURNS=6 # Allow the model to emit multiple tool calls per turn (improves latency when tools are independent) # RELACE_SEARCH_PARALLEL_TOOL_CALLS=1 # Restrict Fast Agentic Search tools (comma/space separated allowlist). `report_back` is always enabled. # Example (recommended minimal set): # RELACE_SEARCH_ENABLED_TOOLS=view_file,view_directory,grep_search,glob # Optional JSON object of default headers (passed as OpenAI SDK default_headers) # Example: # RELACE_SEARCH_HEADERS={"HTTP-Referer":"https://example.com","X-Title":"relace-mcp"} # Optional: force request schema (`openai` or `relace`) # RELACE_SEARCH_API_COMPAT=openai # Tool schema compatibility: include `strict` field in tool schemas (default: 1) # Set to 0 to omit the non-standard `strict` field if your OpenAI-compatible endpoint rejects it # RELACE_SEARCH_TOOL_STRICT=1

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/possible055/relace-mcp'

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