According to AlphaMoat, as of 2026-09, MemoryLayer has 3,562 downloads and 7 stars — ranked #3,072 of 5,078 AI skills overall, and #622 of 969 in AI Agent.
| Platform | Installs | Downloads | Stars | Comments |
|---|---|---|---|---|
| ClawHub ↗ | 86 | 2.7K | 7 | - |
| SkillHub ↗ | 450 | 3.6K | 7 | — |
Platforms count differently and update at different times.
The documentation shows memory storage, meaning-based search, prompt-context retrieval, type and metadata tagging, and usage statistics. Access uses an account or API key, with SDK-style examples. These examples describe an integration workflow; service availability and the supplied SDK interfaces still require verification.
Profile published 2026-10-27
Generated from the skill’s documentation as examples; not results of actual runs.
The counting method changed in 2026-09: skills with the same name are now counted separately by author, and the full catalog is covered. Dashed lines show months before the change; they are not compared with later months, and no growth is computed across the change.
| Month | Downloads | MoM | Stars |
|---|---|---|---|
| 2026-09 | 3.6K | — | 7 |
| 2026-08 | 3.2K | +2.3% | 7 |
| 2026-07 | 3.1K | +5.5% | 7 |
| 2026-06 | 2.9K | +4.1% | 7 |
| 2026-05 | 2.8K | — | 7 |
| 2026-03 | 2.4K | — | 7 |
The author’s original file, shown as plain text without edits or review by AlphaMoat. Links are not clickable; check commands before running them. Fetched 2026-10-23.
---
slug: memorylayer
name: MemoryLayer
description: Semantic memory for AI agents. 95% token savings with vector search.
homepage: https://memorylayer.clawbot.hk
metadata:
clawdbot:
emoji: "🧠"
---
# MemoryLayer Semantic memory infrastructure for AI agents that actually scales. ## Features - **95% Token Savings** - Retrieve only relevant memories - **Semantic Search** - Find memories by meaning, not keywords - **Sub-200ms** - Lightning-fast memory retrieval - **Multi-tenant** - Isolated memory per agent instance ## Setup ### 1. Sign up for FREE account Visit https://memorylayer.clawbot.hk and sign up with Google. You'll get: - 10,000 operations/month - 1GB storage - Community support ### 2. Configure credentials ```bash # Option 1: Email/Password export [email protected] export MEMORYLAYER_PASSWORD=your_password # Option 2: API Key (recommended for production) export MEMORYLAYER_API_KEY=ml_your_api_key_here ``` ### 3. Install Python SDK (if not using skill wrapper) ```bash pip install memorylayer ``` ## Usage ### Basic Example ```javascript // In your Clawdbot agent const memory = require('memorylayer'); // Store a memory await memory.remember( 'User prefers dark mode UI', { type: 'semantic', importance: 0.8 } ); // Search memories const results = await memory.search('UI preferences'); console.log(results[0].content); // "User prefers dark mode UI" ``` ### Python Example ```python from plugins.memorylayer import memory # Store memory.remember( "Boss prefers direct reporting with zero bullshit", memory_type="semantic", importance=0.9 ) # Search results = memory.recall("What are Boss's preferences?") for r in results: print(f"{r.relevance_score:.2f}: {r.memory.content}") ``` ### Token Savings **Before MemoryLayer:** ```python # Inject entire memory files context = open('MEMORY.md').read() # 10,500 tokens prompt = f"{context}\n\nUser: What are my preferences?" ``` **After MemoryLayer:** ```python # Inject only relevant memories context = memory.get_context("user preferences", limit=5) # ~500 tokens prompt = f"{context}\n\nUser: What are my preferences?" ``` **Result:** 95% token reduction, $900/month savings at scale ## API Reference ### `memory.remember(content, options)` Store a new memory. **Parameters:** - `content` (string): Memory content - `options.type` (string): 'episodic' | 'semantic' | 'procedural' - `options.importance` (number): 0.0 to 1.0 - `options.metadata` (object): Additional tags/data **Returns:** Memory object with `id` ### `memory.search(query, limit)` Search memories semantically. **Parameters:** - `query` (string): Search query (natural language) - `limit` (number): Max results (default: 10) **Returns:** Array of SearchResult objects ### `memory.get_context(query, limit)` Get formatted context for prompt injection. **Parameters:** - `query` (string): What context do you need? - `limit` (number): Max memories (default: 5) **Returns:** Formatted string ready for prompt ### `memory.stats()` Get usage statistics. **Returns:** Object with `total_memories`, `memory_types`, `operations_this_month` ## Advanced ### Memory Types **Episodic** - Events and experiences ```javascript memory.remember('Deployed MemoryLayer on 2026-02-03', { type: 'episodic' }); ``` **Semantic** - Facts and knowledge ```javascript memory.remember('Boss prefers concise reports', { type: 'semantic' }); ``` **Procedural** - How-to and processes ```javascript memory.remember('To restart server: ssh root@... && systemctl restart...', { type: 'procedural' }); ``` ### Metadata Tagging ```javascript memory.remember('User likes blue', { type: 'semantic', metadata: { category: 'preferences', subcategory: 'colors', source: 'user_profile' } }); ``` ### Usage Tracking ```javascript const stats = await memory.stats(); console.log(`Total memories: ${stats.total_memories}`); console.log(`Operations this month: ${stats.operations_this_month}`); console.log(`Plan: ${stats.plan} (${stats.operations_limit}/month)`); ``` ## Pricing **FREE Plan** (Current) - 10,000 operations/month - 1GB storage - Community support **Pro Plan** ($99/mo) - 1M operations/month - 10GB storage - Email support - 99.9% SLA **Enterprise** (Custom) - Unlimited operations - Unlimited storage - Dedicated support - Self-hosted option - Custom SLA ## Support - **Documentation:** https://memorylayer.clawbot.hk/docs - **API Reference:** https://memorylayer.clawbot.hk/api - **Community:** Discord (link in docs) - **Issues:** GitHub (link in docs) ## Links - **Homepage:** https://memorylayer.clawbot.hk - **Dashboard:** https://dashboard.memorylayer.clawbot.hk - **API Docs:** https://memorylayer.clawbot.hk/docs - **Python SDK:** https://pypi.org/project/memorylayer (when published)
Records learnings, errors and corrections as Markdown files for later reuse and promotion into workspace memory.
Finds skills in local and remote catalogs from a plain-language request and recommends what to install.
A browser-automation command line skill for AI agents.
Turns corrections and self-review into tiered memory and promotes or archives it by use.
Reviews a third-party skill's source, permissions, and danger patterns against a checklist before it is installed.
Stores agent memory and cross-skill state as a typed, constraint-checked knowledge graph.
A workspace setup guide for planning proactive follow-up, working memory and review routines for an AI agent.
A guide to creating and updating effective Claude skills.
Data month: 2026-09 · Downloads, stars and installs are aggregated monthly from public skill registries (ClawHub, SkillHub). See methodology.