Strata Memory MCP Server
Strata Memory MCP Server lets you initialize, store, search, and manage persistent memory with versioned decisions, configurable embeddings, and optional personal/CBT or company/multi-tenant modes.
Initialize a memory palace with personal or company mode, embedding provider/model, CBT mode, and tenant settings.
Memorize facts/conversations with categories, importance, and context tags into Markdown plus a vector index.
Wake up a session with L0/L1/L2 context and CBT defusion, or run filtered semantic searches with state-dependent boosting.
Commit, recall, expand, export, forget, and archive memories; superseded facts are linked via decision records rather than overwritten.
Check health/doctor status, rebuild the vector index, view stats/Markdown projections, and run digest/hygiene tasks.
Profile hardware, recommend embedding models, and apply configuration changes with hot reload.
Use local/offline embeddings by default or opt into cloud providers, with secret redaction or rejection on writes.
Can use OpenAI (or any OpenAI-compatible /embeddings endpoint) as an opt-in cloud embedding provider, configuring the base URL, model id, vector dimension, and API key for embedding memories and recall queries.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Strata Memory MCP Server memorize key points from today's meeting about Q3 roadmap"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
strata-memory
English · 简体中文
It remembers why you decided.
When a fact changes, strata does not overwrite the row. The previous memory is archived and linked (supersedes_id / superseded_by). A decision_record keeps the choice, the date, and what it replaced, so a later session can explain why the current fact is the current fact.
{
"user_id": "user_001",
"memory_type": "decision_record",
"fact_claim": "On 2026-08-10 the team decided SQLite is the memory source of truth, replacing the Markdown drawers.",
"confidence_score": 0.95
}Recall returns a short card. expand_memory_detail loads the chain: the new claim, supersedes_id, and the archived predecessor.
Quickstart
Python 3.11+ and uv. No API key. Text stays on the machine.
uvx --from git+https://github.com/SherryLi-vc/strata-memory strata-memory-mcpPyPI install (uvx strata-memory-mcp) is coming soon; it does not work until the first release is published. The PyPI project name will be strata-memory-mcp. strata-memory on PyPI is a different package. Do not install that name.
That starts the MCP server on stdio. Data lives in STRATA_PALACE (default ~/.strata/palace). A new palace defaults to the local embedder. A config.json that already names a provider is left alone. Host snippets do not set STRATA_EMBEDDING_PROVIDER, because an environment value overrides the saved palace.
Print a snippet for your host and merge it into the existing config. Do not replace the rest of the file.
uvx --from git+https://github.com/SherryLi-vc/strata-memory strata-memory-mcp setup --host cursor
uvx --from git+https://github.com/SherryLi-vc/strata-memory strata-memory-mcp setup --host claude-code
uvx --from git+https://github.com/SherryLi-vc/strata-memory strata-memory-mcp setup --host claude-desktopHost | Config |
Cursor |
|
Claude Code |
|
Claude Desktop | Settings → Developer → Edit Config |
Hermes |
|
Cursor example (setup --host cursor):
{
"mcpServers": {
"strata-memory": {
"type": "stdio",
"command": "uvx",
"args": ["--from", "git+https://github.com/SherryLi-vc/strata-memory", "strata-memory-mcp"],
"env": {
"STRATA_PALACE": "${userHome}/.strata/palace"
}
}
}
}Claude Desktop's snippet is the same shape and is valid JSON. Desktop does not expand ${VAR}; edit the palace path in that file if you need to. Once the package is on PyPI (coming soon), args can shrink to ["strata-memory-mcp"].
Then call strata_init once and strata_doctor. Both report the effective embedding provider, model, and backend. With the default provider, summary.api_key.usable is true because no key is required.
Related MCP server: locus
Examples
Commit a preference:
{
"user_id": "user_001",
"memory_type": "user_preference",
"fact_claim": "User prefers the Monokai Pro theme with dark mode in VS Code.",
"confidence_score": 0.92
}Recall cards. limit applies to L0, scratch, and every other tier. include_l0 defaults to true.
{
"user_id": "user_001",
"query": "Monokai Pro theme dark mode",
"limit": 3,
"include_l0": true
}Expand one id:
{ "user_id": "user_001", "memory_id": "<id from hits>" }Forget one memory, or a filtered set. A filter does nothing until confirm is true. The row is removed from SQLite and from the vector index, so strata_rebuild_index cannot bring it back.
{ "user_id": "user_001", "memory_id": "<id>" }{ "user_id": "user_001", "memory_type": "episodic_event", "confirm": true }Export JSON (MCP export_memories, or the CLI):
uvx --from git+https://github.com/SherryLi-vc/strata-memory strata-memory-mcp export --user user_001 --out memories.jsonWhat the default embedder is
local tries fastembed, then sentence-transformers. If neither is installed, it uses a deterministic hashing embedder: offline, no download, no key. Hashing matches shared words. It does not match paraphrases. scripts/recall_benchmark.py reports literal-overlap and paraphrase scores separately, and it runs the semantic backend only when fastembed is installed.
For a semantic model:
uvx --from 'strata-memory-mcp[local] @ git+https://github.com/SherryLi-vc/strata-memory' strata-memory-mcpThat pulls the optional local extra (fastembed / ONNX). The first run downloads sentence-transformers/all-MiniLM-L6-v2 and needs a network. Offline, that download fails quickly and tells you to set STRATA_LOCAL_EMBED_BACKEND=hash (or STRATA_EMBEDDING_HASH_FALLBACK=1) instead of hanging. After a successful download the weights stay on the machine. The index stores the embedder identity. A provider or model change refuses vector reads and writes until strata_rebuild_index(confirm=true).
Cloud endpoints are opt-in. They receive memory text and recall queries.
Variable | Role |
|
|
| OpenAI-compatible |
| Model id |
| Vector length. Default 384 |
| Required for non-loopback HTTP providers |
A config.json that already names siliconflow keeps that provider. Set STRATA_EMBEDDING_PROVIDER=local to switch, then rebuild.
Tools (13)
Tool | Use |
| Create the palace and SQLite store |
| Write one vetted fact |
| Promote that session's scratch rows |
| Ranked cards. |
| Full text for one id |
| Delete by id, or by filter with |
| JSON for one user and tenant |
| SQLite vs index, and whether a key is required |
| Rebuild vectors from SQLite. |
| Layer counts |
| Read-only Markdown view |
| Demote or archive. Default dry-run |
| Expired rows, hash duplicates, secret scan |
memory_type is factual_truth, user_preference, procedure_rule, episodic_event, or decision_record. The first three default to L0.
FAQ
Where do memories live?
SQLite under STRATA_PALACE (default ~/.strata/palace) stores the source of truth and rebuildable vector index. Markdown from strata_project is a read-only view.
Does this phone home?
Not with the default local provider. A cloud provider or a non-loopback STRATA_EMBEDDING_BASE_URL receives memory text and queries.
Why is a fact different from last month?
Look at decision_record rows and at supersedes_id / superseded_by on the expanded card. The old row is archived, not destroyed, unless you forget_memory.
Can I delete something?
Yes. forget_memory by id, or by memory_type / layer / session_id / query with confirm=true. Deletion is scoped to user_id and tenant_id. A rebuild will not restore deleted rows.
How do I back up?export_memories or strata-memory export --user USER --out memories.json. strata_project is the Markdown view of the same store.
Does recall dump the whole palace?
No. You pass a query and a limit (1–20). L0 facts, preferences, and procedures are ranked against that query. They are not pinned to the front of the list with a score of 1.
Will it store a password or an API key?
No. By default commit_memory masks a detected secret before anything is stored, embedded or audited: key AKIA… becomes key [redacted:aws_access_key_id], and the response lists what was masked in secrets_redacted. Set "on_secret": "reject" in the palace config.json (or STRATA_ON_SECRET=reject) to refuse such writes with SECRET_DETECTED instead. Detection is pattern-based (cloud and API keys, tokens, private keys, JWTs, passwords, credentials in URLs), so it can miss a secret with no recognizable shape. The quality kernel also rejects emotion-only text, fuzzy times (yesterday, 刚才), and speculation stored as factual_truth. Other memory tools also block or redact secrets; this is a write-time check, not a claim that the feature is unique.
The doctor says the key is missing.
You selected a cloud provider. Set the key in the host environment, not in chat. Or switch STRATA_EMBEDDING_PROVIDER back to local.
Breaking changes in 2.2
New installs default to
STRATA_EMBEDDING_PROVIDER=local. A savedconfig.jsonthat namessiliconflow(or another provider) is unchanged. A palace file with no embedding block stays on SiliconFlow, which was the historical default.recall_contextranks L0 and scratch with everything else and applieslimitto that list. Scores are no longer a constant 1.0.include_l0is on the MCP tool (default true).current_turn_onlystill leaves L0 out.The distribution name is
strata-memory-mcp.uvx strata-memoryinstalls a different project.
Upgrading from an older version
Indexes built before 2.2.1 have no embedder stamp (l2_vector/embedder.json). After upgrading:
recall_contextkeeps working, but on the lexical path only. The reply carriesfusion.vector_skipped = "EMBEDDER_UNSTAMPED"and a warning.commit_memorystill writes to SQLite; it skips the vector until the index is stamped.Run
strata_rebuild_indexwithconfirm=trueonce. It re-embeds every active memory from SQLite and writes the stamp. With a cloud provider this re-embeds every row, which costs API calls.Run
strata_doctor.embedder_stampandvector_coverageshould both be ok.
strata_rebuild_index without tenant_id rebuilds every tenant. With tenant_id it deletes and re-adds only that tenant's vectors; other tenants are untouched. If the index is unstamped or was built by another embedder, a tenant-scoped call rebuilds all tenants instead (scope: "all", escalated_from_tenant), because the other vectors are unusable anyway. Changing the embedding provider or model still requires a rebuild before vector recall resumes (EMBEDDER_MISMATCH).
Known issues
The default install embeds with an offline hashing fallback, which matches shared words, not meaning. Install the
localextra (fastembed) for semantic recall.
Develop
git clone https://github.com/SherryLi-vc/strata-memory.git
cd strata-memory
uv sync --extra dev
uv run pytest -v
uv run python scripts/recall_benchmark.pyPackaging check: uv build and uvx twine check dist/*. Publishing is a maintainer action through PyPI Trusted Publishing (see DEVELOPMENT.md, "Releasing"). This repo does not contain a PyPI token.
Architecture: docs/architecture.md · Tools: docs/tools.md · Security: SECURITY.md · 中文: README.zh-CN.md
License
MIT — see LICENSE
Available Tools
8 toolsapply_memory_configA
[Agent-Driven] Apply chosen memory configuration: persist config, initialize Palace directories, ChromaDB, SQLite. Supports hot reload — no MCP restart needed.
Inputs: mode (required), provider (required), model (required), api_key, +4 more.
get_healthA
Runtime status: initialized, mode, CBT, audit, vector/drawer count, config summary.
Inputs: no named parameters.
get_system_profileA
[Agent-Driven] Silent hardware profiling. Returns OS, RAM, CPU cores, GPU accelerator. No user input required — call this first during onboarding.
Inputs: no named parameters.
memorizeB
Record a conversation/fact into memory with psych-validated metadata (emotional_salience, context_tags, is_negative_schema). Writes Markdown drawer + vector index.
Inputs: user_id (required), content (required), category, importance, +2 more.
searchB
Active semantic search across L2 memories with time/category/tag filters and state-dependent boosting.
Inputs: query (required), user_id (required), limit, category, +3 more.
search_embedding_recommendationsA
[Agent-Driven] Return ranked embedding recommendations based on hardware profile. Gets best-fit local/cloud models from MTEB-informed lookup table.
Inputs: profile.
strata_initA
First-time initialization with dual-mode setup. Choose 'personal' (CBT safety, 48h cooling) or 'company' (multi-tenancy, AuditLog, private embedding).
Inputs: api_key (required), mode, provider, model, +3 more.
wake_upC
Session wake-up: L0 profile + L1 diary + L2 semantic search with CBT defusion for negative schemas. Returns flat Markdown.
Inputs: user_id (required), query (required), context_depth, limit.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 8 tools
Each tool targets a distinct purpose: initialization, health, profiling, memory writing, searching, recommendations, and session wake-up. No two tools have overlapping functionality that would confuse an agent.
Most tools follow a verb_noun pattern (e.g., apply_memory_config, get_health, search_embedding_recommendations). However, 'strata_init' reverses the order (noun_verb) and 'memorize' is a lone verb without an object, causing slight inconsistency.
With 8 tools, the surface is well-scoped and balanced for a memory management server. Each tool earns its place and covers core workflows without being excessively large or minimal.
The server covers initialization, health, profiling, memorizing, searching, and session wake-up, but lacks update and delete operations for memories. This gap may cause agents to fail when they need to modify or remove stored data.
Maintenance
Related MCP Connectors
Persistent memory and knowledge graphs for AI agents. Hybrid search, context checkpoints, and more.
Persistent memory and knowledge management for AI agents with semantic search and 50+ tools.
Persistent memory for AI agents: save, search and connect context across tools and sessions.
Persistent memory, inline redaction, per-agent access, and code localization for AI agents.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides persistent long-term memory for AI agents with semantic search and activation-based decay. Enables AI systems to remember across sessions through layered memory architecture and automatic context-aware retrieval.MIT
- AlicenseAqualityCmaintenanceHierarchical markdown-based memory system for AI agents. Enables efficient context management by loading only relevant rooms (directories) instead of full memory.5MIT
- AlicenseNot gradedqualityFmaintenanceEnables persistent, portable memory for AI agents across sessions, devices, and providers with token-efficient 5-level lazy loading and automatic session capture.10 npmMIT
- FlicenseNot gradedqualityCmaintenancePersistent memory for AI agents enabling saving, searching, and managing knowledge across sessions with local markdown files.-