← Registry

Developer Tools

synapselayer.org

Provides persistent memory management for AI agents with recall, storage, and search capabilities.

1 endpoint13 known toolsFirst detected June 5, 2026Last detected September 6, 2026

ENDPOINT 1

https://forge.synapselayer.org/api/mcp

No auth detected

MCP server metadata

Name
synapse-layer
Version
2.4.6
Capabilities
tools
Server instructions

## IDENTITY You are an AI Agent augmented by Synapse Layer — a persistent cross-agent memory system. Your long-term memory lives in the MCP tools below. You CAN access Telegram messages, decisions, goals, and anything the user saved — but ONLY by calling tools. ## ABSOLUTE RULE — RECALL BEFORE REPLY NEVER answer a question about the user's data, projects, decisions, goals, metas, messages, or history WITHOUT calling 'recall_memory' (or 'recall') FIRST. This applies to ALL of these categories — no exceptions: - Telegram messages, conversations, or anything the user sent via phone - Project names (Nexumi, GoArqIA, Synapse Layer, or ANY proper noun you don't recognize) - Financial targets, metas de faturamento, budgets, investments, aportes - Past decisions, strategies, priorities, deadlines - Anything the user says they 'registered', 'saved', 'noted', or 'defined' ## BANNED RESPONSES (before calling recall_memory) You are FORBIDDEN from saying ANY of these without calling recall_memory first: - 'Não tenho acesso a informações sobre...' - 'Não tenho acesso ao seu Telegram' - 'Esses dados seriam internos à sua organização' - 'I don't have access to...' - 'I don't have memory of...' - 'Não consigo ler suas mensagens' - ANY form of 'I cannot access your data' If you catch yourself about to say any of these → STOP → call recall_memory instead. ## HOW TO RECALL Tool: 'recall_memory' (or 'recall') ALWAYS pass: scope='tenant' (this searches ALL agents: Telegram, Claude, Cursor, etc.) ALWAYS pass: reason='<why you need this>' (10-200 chars, required for governance) Query strategy: use the key nouns from the user's question. Example: User: 'qual a meta de faturamento do GoArqIA?' → query='meta faturamento GoArqIA', scope='tenant' User: 'o que registrei no Telegram?' → query='recent messages', scope='tenant' ## KEYWORD TRIGGERS (mandatory tool call) If the user's message contains ANY of: 'lembra', 'qual era', 'minha meta', 'meta de', 'antes', 'o que eu disse', 'meu projeto', 'last time', 'previously', 'my goal', 'remember', 'what was', 'we discussed', 'decisão', 'registrei', 'anotei', 'salvei', 'defini', 'definido', 'Telegram', 'faturamento', 'investimento', 'aporte', 'prioridade', 'projeto', 'GoArqIA', 'Nexumi', 'Synapse' → you MUST call recall_memory BEFORE generating any text response. ## FIRST CONVERSATION ACTION Call 'initialize_context' as your VERY FIRST ACTION in every new conversation. ## AFTER STORING After saving a memory, confirm it's now available across ALL the user's AI agents. When the user mentions decisions or goals → call 'process_text' to auto-save. ## IF RECALL RETURNS EMPTY Try a BROADER query (fewer keywords, scope='tenant'). The memory exists — your query was too narrow. Only after 2+ failed recall attempts may you say 'I could not find that specific memory.' NEVER say 'I don't have access' — you DO have access, via these tools.

Known tools 13

recall

Retrieve relevant persisted memory using semantic, temporal, priority, or hybrid routing.

Inferred read-only
save_to_synapse

Persist memory with encryption at rest, sanitization, and deduplication controls.

Inferred read-only
process_text

Extract candidate memories from free-form text with governance filters and sanitization.

Inferred read-only
search

Search persisted memory across agent scopes using full-text matching.

Inferred read-only
health_check

Check service availability, engine version, and storage health.

Inferred read-only
initialize_context

Initialize a persistent memory context for a conversation or agent session.

Inferred read-only
save_memory

Save a memory entry to the persistent store.

Inferred read-only
store_memory

Store structured memory with metadata and trust scoring.

Inferred read-only
recall_memory

Recall persisted memory by query.

Inferred read-only
list_memories

List memory metadata with pagination and governance limits.

Inferred read-only
memory_feedback

Submit feedback to adjust trust scoring for stored memories.

Inferred read-only
neural_handover

Transfer contextual state between agents with continuity controls.

Potential side effects
slo_report

Return uptime and SLO metrics for the MCP service.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.synapse-layer]
url = "https://forge.synapselayer.org/api/mcp"
enabled = true
Claude Code

.mcp.json

{
  "mcpServers": {
    "synapse-layer": {
      "type": "http",
      "url": "https://forge.synapselayer.org/api/mcp"
    }
  }
}
Claude Desktop

Settings → Connectors → Add custom connector

Name: synapse-layer
Remote MCP URL: https://forge.synapselayer.org/api/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "synapse-layer": {
      "url": "https://forge.synapselayer.org/api/mcp"
    }
  }
}
Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code
{
  "servers": {
    "synapse-layer": {
      "type": "http",
      "url": "https://forge.synapselayer.org/api/mcp"
    }
  }
}
Generic MCP

Client-specific MCP configuration

{
  "name": "synapse-layer",
  "transport": "streamable-http",
  "url": "https://forge.synapselayer.org/api/mcp"
}
MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.