MCP.Market

Hot Memory MCP server

by michael-denyer·io.github.michael-denyer/hot-memory-mcp·v0.8.0·9 stars
No reviews yet

Two-tier memory: hot cache (0ms) + semantic search. Self-organizing.

A92/100grade A
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
A92/100

full report

Adoption
Growing

9 stars201 downloads/wk

Reviews

Write one

Nobody has reviewed Hot Memory yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Hot Memory tools (56, 8 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • access_patterns

    Get learned access patterns for predictive caching.

  • approve_candidate

    Approve a mined pattern, storing it as a memory and optionally promoting to hot cache.

  • audit_history

    Get recent audit log entries for destructive operations.

  • bootstrap_project

    Bootstrap hot cache from project documentation files.

  • bulk_reject_candidates

    Bulk reject multiple mining candidates at once.

  • cross_session_patterns

    Find content patterns appearing across multiple conversation sessions.

  • db_info

    Get database information including schema version and size.

  • db_maintenancewrite action

    Run database maintenance (vacuum, analyze, auto-demote stale).

  • db_rebuild_vectors

    Rebuild all memory vectors with the current embedding model.

  • demotewrite action

    Remove a memory from hot cache (keeps in cold storage).

  • embedding_info

    Get embedding provider and cache information.

  • end_session

    End a session and consolidate episodic memories.

  • find_contradictions

    Find memories that may contradict a given memory.

  • forgetwrite action

    Delete a memory permanently.

  • get_contradictions

    Get all memory pairs marked as contradictions.

  • get_related_memories

    Get memories related to a given memory.

  • get_session

    Get details for a specific session.

  • get_session_memories

    Get all memories from a specific conversation session.

  • get_sessions

    Get recent conversation sessions.

  • get_trust_history

    Get the trust adjustment history for a memory.

  • hot_cache_status

    Show current hot cache contents, stats, and observability metrics.

  • invalidate_memory

    Mark a memory as incorrect or outdated.

  • link_memorieswrite action

    Create a typed relationship between two memories.

  • list_memories

    List stored memories with pagination.

  • log_output

    Log an output for pattern mining. Called automatically or manually.

  • mark_contradiction

    Mark two memories as contradicting each other.

  • mark_memory_used

    Mark a memory as actually used/helpful after recall.

  • memory_stats

    Get memory statistics.

  • metrics_status

    Get observability metrics for monitoring and debugging.

  • mining_status

    Show pattern mining statistics.

  • pin

    Pin a hot cache memory to prevent auto-eviction.

  • predict_next

    Predict which memories might be needed next based on access patterns.

  • predictive_cache_status

    Get status of the predictive hot cache system.

  • preview_consolidation

    Preview memory consolidation without making changes.

  • promote

    Manually promote a memory to hot cache for instant recall.

  • recall

    Semantic search with confidence gating and composite ranking.

  • recall_by_tag

    Get memories with a specific tag.

  • recall_with_fallback

    Recall with automatic fallback through memory types.

  • reject_candidate

    Reject a mined pattern, removing it from candidates.

  • relationship_stats

    Get statistics about memory relationships in the knowledge graph.

  • remember

    Store a new memory. Returns the memory ID.

  • resolve_contradiction

    Resolve a contradiction by keeping one memory and handling the other.

  • retrieval_quality_stats

    Get retrieval quality statistics.

  • review_candidates

    Review mined patterns that are ready for promotion.

  • run_cleanupwrite action

    Run comprehensive cleanup of stale data.

  • run_consolidationwrite action

    Consolidate similar memories by merging near-duplicates.

  • run_miningwrite action

    Run pattern mining on recent output logs.

  • seed_from_file

    Seed memories from a file.

  • seed_from_text

    Seed memories from text content.

  • set_session_topic

    Set or update the topic for a conversation session.

  • summarize_session

    Summarize a session's key decisions, insights, and action items.

  • unlink_memorieswrite action

    Remove relationship(s) between two memories.

  • unpin

    Unpin a memory, making it eligible for auto-eviction from hot cache.

  • validate_embeddings

    Check if the embedding model has changed since database was created.

  • validate_memory

    Mark a memory as validated/confirmed useful.

  • warm_cache

    Pre-warm hot cache with predicted next memories.

Public scan report

scanner v0.1.9 · 2026-09-26 · same rubric, same numbers if you re-run it

no findings
  • Code scan58 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 10 days ago15/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 92/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Hot Memory compares

Against every graded Web search MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
92/100
higher than 90% of 3,470 Web search servers
Tools
56
the typical Web search server has 5
Tools that change things
14%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
9
more than 85% of 1,786 Web search servers
Downloads a week
201
more than 70% of 827 Web search servers with a package

What its tools let an agent do: 44 read · 9 change · 3 run or deploy

Install Hot Memory in Claude Code, Cursor or VS Code

claude mcp add hot-memory-mcp -- uvx hot-memory-mcp
Add to Cursor

What the publisher says

From the Hot Memory repository's README, as published. We do not edit it. Read it on GitHub

<!-- mcp-name: io.github.michael-denyer/hot-memory-mcp -->

🧠 Memory MCP

Give your AI assistant a persistent second brain

Stop re-explaining your project every session.

Memory MCP learns what matters and keeps it ready — instant recall for the stuff you use most, semantic search for everything else.

The Problem

Every new chat starts from scratch. You explain your architecture again. You paste the same patterns again. Your context window bloats with repetition.

Other memory solutions help, but they still require tool calls for every lookup — adding latency and eating into Claude's thinking budget.

Memory MCP fixes this with a two-tier architecture:

  1. Hot cache (0ms) — Frequently-used knowledge printed into context by the plugin's SessionStart and UserPromptSubmit hooks before Claude even starts thinking. Claude Code adds a hook's stdout to the conversation, so no tool call is needed.
  2. Cold storage (~50ms) — Everything else, searchable by meaning via semantic similarity.

The system learns what you use and promotes it automatically. Your most valuable knowledge becomes instantly available. No manual curation required.

Before & After

Install

# Install package
uv tool install hot-memory-mcp   # or: pip install hot-memory-mcp

# Add plugin (recommended)
claude plugins add michael-denyer/memory-mcp

The plugin gives you auto-configured hooks and slash commands. Embeddings run on sentence-transformers on every platform, including Apple Silicon.

Manual config (no plugin)

Add to ~/.claude.json:

{
  "mcpServers": {
    "memory": {
      "command": "memory-mcp"
    }
  }
}

See Reference for full configuration options.

Restart Claude Code. Run memory-mcp-cli bootstrap once to seed memories from your project docs; the hooks inject whatever is promoted, but nothing seeds itself.

First run: Embedding model (~90MB) downloads automatically. Takes 30-60 seconds once.

How It Works

flowchart LR
    subgraph LLM["Claude"]
        REQ((Request))
    end

    subgraph Hot["HOT CACHE · 0ms"]
        HC[Session context]
        PM[(Promoted memories)]
    end

    subgraph Cold["COLD STORAGE · ~50ms"]
        VS[(Vector search)]
        KG[(Knowledge graph)]
    end

    REQ -->|"hook stdout"| HC
    HC -.->|"draws from"| PM
    REQ -->|"recall()"| VS
    VS <-->|"related"| KG

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Hot Memory: common questions

Is Hot Memory MCP server safe?
Yes, by our scan: it is graded A (92/100). Read the Hot Memory safety report
How do I install Hot Memory?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Hot Memory need an API key?
No secret keys are declared. It reads 2 settings from the environment.
Is Hot Memory maintained?
The last commit was 49 days ago (2026-09-16). The latest release is v0.8.0.
What can I use instead of Hot Memory?
Servers from other publishers that do the same job: Marm MCP server, Shodh Memory MCP server and Octobrain MCP server. Compare all Hot Memory alternatives.
How does Hot Memory compare with other Web search MCP servers?
safety score 92/100 (higher than 90% of 3,470 Web search servers); tools 56 (the typical Web search server has 5); tools that change things 14% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to Hot Memory

Same job from other publishers: the closest match first, then the best rated.

All Hot Memory alternatives →
  • Marm
    Universal MCP Server with advanced AI memory capabilities and semantic search.
    A
  • Shodh Memory
    Cognitive memory for AI agents — semantic search, Hebbian learning, knowledge graphs.
    C
  • Octobrain
    Persistent memory for AI assistants with semantic search and knowledge graph relationships.
    More from Muvon →
    A
  • Arxiv
    Search arXiv papers, download full text, semantic search, citation graphs, and alerts via MCP.
    A
  • Socraticode
    MCP server for enterprise local codebase indexing, semantic search, and code dependency graphs.
    B