Agentic memory for CTI in Python — STIX knowledge graphs, threat-actor alias resolution, offline-first RAG, MCP server for Claude Code and LangChain agents
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add zettelforge -- uvx zettelforge~/.codex/config.toml
[mcp_servers.zettelforge]
command = "uvx"
args = ["zettelforge"]
# codex mcp add zettelforge -- uvx zettelforge~/.cursor/mcp.json
{
"mcpServers": {
"zettelforge": {
"command": "uvx",
"args": [
"zettelforge"
]
}
}
}.vscode/mcp.json
{
"servers": {
"zettelforge": {
"type": "stdio",
"command": "uvx",
"args": [
"zettelforge"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"zettelforge": {
"command": "uvx",
"args": [
"zettelforge"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"zettelforge": {
"command": "uvx",
"args": [
"zettelforge"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"zettelforge": {
"command": "uvx",
"args": [
"zettelforge"
]
}
}
}