Gnosem
by gnosem.dev · remote · no sign-in to list tools · v1.0.0 · published by dev.gnosem
Cross-vendor AI memory over MCP. One semantic store, readable and writeable from every MCP client.
6 tools: 4 change things and 2 only read.
Scanned Oct 21, 2026
- Deletes or cancels0Claude Code: deny
- Can't be undone0Claude Code: ask
- Changes things4Claude Code: ask: see these tools
- Only reads2Claude Code: allow: see these tools
No check matched anything in the tool definitions. No finding is not a safety guarantee. How the checks work.
claude mcp add --transport http gnosem https://gnosem.dev/mcpTools
Open a tool for its reasons and full description. Descriptions are the publisher's own words, shown as they are.
6 tools
Changes things 4Claude Code: ask
memory_forgetSoft-delete a memory by id.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Soft-delete a memory by id. The row is retained for audit but excluded from search/list and removed from the vector index.
memory_supersedeReplace a stale memory with a corrected one.No findings
Why it is in this group
- The description's first verb is "replace": a word for a change.
Description, as the server wrote it Replace a stale memory with a corrected one. The old row is marked superseded and excluded from future reads; the new row becomes the current version. Use for corrections; use memory_forget for pure deletions.
memory_writeSave a fact, preference, decision, or note to the user's cross-model memory.No findings
Why it is in this group
- The name "memory_write" says "write": a word for a change.
Description, as the server wrote it Save a fact, preference, decision, or note to the user's cross-model memory. Any MCP client can read this back later. Include written_by (e.g. 'claude-code', 'gpt-5', 'kimi-k2') for provenance and session_id to group related writes. Long content (>400 chars) is automatically compressed on write to a structured-facts form optimized for LLM reading — the raw text is preserved. Pass no_optimize:true to skip. Writes are deduped by default: (1) SHA-256 of trim(content) short-circuits byte-identical writes with { id, exact_duplicate:true } for free (no embed call); (2) failing that, semantic dedup returns { id, deduped:true, matched_score } when cosine ≥ 0.85. Pass force:true to bypass both, or use memory_supersede to explicitly correct a prior memory.
memory_write_bulkWrite up to 50 memories in a single call.No findings
Why it is in this group
- The name "memory_write_bulk" says "write": a word for a change.
Description, as the server wrote it Write up to 50 memories in a single call. Each entry runs the same path as memory_write (semantic dedup by default; pass force:true per-entry to skip). Embeddings + optimizations run in parallel; D1 inserts are batched. Returns { results: [...] } with one entry per input in the same order — each is { id, created_at, optimized? } on success, { id, created_at, deduped, matched_score } on dedup, or { error } on failure. Free-tier limits apply to the sum: if adding N would exceed 200, the first (200 - existing) succeed and the rest return an error.
Only reads 2Claude Code: allow
memory_listList the user's most recent memories in reverse chronological order.No findings
Why it is in this group
- The name "memory_list" says "list": a word for reading.
Description, as the server wrote it List the user's most recent memories in reverse chronological order. Use for browsing or catching up on what the user's other model sessions have written recently. Same content/content_raw shape as memory_search. Optional filters (tags, written_by, session_id, since, until) narrow the listing at the SQL level.
memory_searchSearch the user's memories.No findings
Why it is in this group
- The name "memory_search" says "search": a word for reading.
Description, as the server wrote it Search the user's memories. Default mode is 'hybrid': blends semantic (cosine over Vectorize) and keyword (BM25 over SQLite FTS5) hits via Reciprocal Rank Fusion (k=60). Semantic catches paraphrases; keyword catches exact-string hits (IDs, dates, code snippets). Pass mode:'semantic' or mode:'keyword' to run just one. Content defaults to the LLM-optimized (compressed) form when available (raw:true to invert). Excludes forgotten + superseded. Optional filters narrow after retrieval: tags (AND), written_by, session_id, and/or since/until (ms epoch).
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Install
Pinned to version 1.0.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http gnosem https://gnosem.dev/mcpRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__gnosem__memory_search",
"mcp__gnosem__memory_list"
],
"ask": [
"mcp__gnosem__memory_write",
"mcp__gnosem__memory_forget",
"mcp__gnosem__memory_supersede",
"mcp__gnosem__memory_write_bulk"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. It assumes the server is installed as gnosem. Review it before you use it.