state-trace
Quality Score
85
/100
Bounded working memory for coding agents — typed nodes, current-vs-stale state, MCP-mountable.
§01 Install
Claude Desktop (uvx)
{
"mcpServers": {
"state-trace": {
"command": "uvx",
"args": [
"state-trace"
],
"env": {
"STATE_TRACE_STORAGE_PATH": "<state_trace_storage_path>",
"STATE_TRACE_NAMESPACE": "<state_trace_namespace>",
"STATE_TRACE_CAPACITY_LIMIT": "<state_trace_capacity_limit>"
}
}
}
}§02 Environment variables
STATE_TRACE_STORAGE_PATHPath to the SQLite or JSON file backing the working-memory graph. .db/.sqlite/.sqlite3 = SQLite (recommended); other = JSON. Defaults to ~/.state-trace/memory.db.
STATE_TRACE_NAMESPACENamespace to scope reads and writes (e.g. the repo slug). Nodes without a namespace remain visible to all namespaces.
STATE_TRACE_CAPACITY_LIMITWorking-memory budget in capacity units. Default 256.
§03 MCP Quality Score · methodology
freshness
25
completeness
15
installability
25
documentation
15
stability
5
§04 Alternatives in Memory & RAG
DoneThat
ai.donethat/donethat
Privacy-first work tracking with summaries, reports, coaching, and AI-ready long-term memory.
Fodda Knowledge Graphs
ai.fodda/mcp-server
Expert-curated knowledge graphs for AI agents — PSFK Retail, Beauty, Sports and more.
ai.meminal/meminal
ai.meminal/meminal
Memory for deep conversational context across any platform