mempalace · Issues· 728 open

chromadb_rust_bindings segfaults on macOS 26.4 (Darwin 25.4) ARM64 — all >=1.5.4,<2 versions affected

mcp_server._get_client has no own-write re-stamp: every add_drawer and check_duplicate reloads the HNSW index (#2307 covered only ChromaBackend)

Global _HTTP_REQUEST_LOCK serializes all requests; a slow read blocks concurrent calls (HTTP transport)

Trust model for multi-agent shared palaces: 3 lessons from actually running one

Wheels for all supported Python versions

mcp_server parses sys.argv at import time: mempalace-light-mcp --help shows the full server's options, and any importer's argv is consumed

kg: invalidate() should symmetrically reject ended < valid_from for the target row

feat: Visual Memory Dashboard for inspecting MemPalace structure and retrieval context

Concurrent writers (hooks + MCP server + CLI) corrupt the palace on chromadb 1.5.8 — sparse-file bloat + SIGSEGV

MCP tool_search returns "Error finding id" when wing-scoped to a convos-mined wing (CLI works, unscoped MCP works)

HNSW pickle corruption after long-running mine: 'unsupported opcode' on next read; repair/migrate cannot recover

mempalace.mcp_server has no single-instance guard — concurrent client spawns race on the palace and trigger HNSW corruption (defense-in-depth for #976)

mempalace-native: read-only open fails on a WAL database without sidecars

mempalace-native: default --db path ignores the XDG config directory

`repair-status` reports `unknown` for healthy palaces with fewer records than `sync_threshold`