// SCAN REPORT

Recall

8 tools. 4 can modify or destroy data without limits.

1 destructive tool with no built-in limits. Policy required.

Last updated: 2 Jul 2026

4 can modify or destroy data

4 read-only

8 tools total

Community server · catalogue entry verified 02/07/2026

How to control Recall ↓

TOOL MAP

What Recall exposes to your agents

Read (4) Write / Execute (3) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Recall tools

Destructive · High forget The forget tool performs irreversible deletion of memory records. Execute · Medium consolidate This tool executes internal algorithms that reorganize or compress the graph memory layer. Execute · Medium bounded_answer This tool executes a generation operation using the Recall memory graph to produce bounded answers. Write · Medium add_memory The 'add_memory' operation creates new memory entries in a typed-edge graph that serves as a persistent knowledge store.

4 of Recall's 8 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Recall

PolicyLayer is an MCP gateway — it sits between your AI agents and Recall, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "forget": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "add_memory": {
    "limits": [
      {
        "counter": "add_memory_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "audit": {
    "limits": [
      {
        "counter": "audit_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Recall — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON RECALL →

Instant setup, no code required.

FULL CATALOGUE

All 8 Recall tools

DESTRUCTIVE 1 tools

Destructive forget Surgically delete a memory by node_id. Audit log preserved.

EXECUTE 2 tools

Execute bounded_answer Generate an answer bounded by Recall Execute consolidate Run sleep-time consolidation now: BMRS pruning + mean-field +

WRITE 1 tools

Write add_memory Add a memory to Recall

READ 4 tools

Read audit Return provenance trail of operations for a target id (or all). Read graph_health Compute graph-theoretic health metrics: spectral gap (λ_2), Read search_memory Search Recall Read stats One-liner stats: active nodes, edges, audit entries.

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Recall

Can an AI agent delete data through the Recall MCP server? +

Yes. The Recall server exposes 1 destructive tools including forget. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Recall? +

The Recall server has 1 write tools including add_memory. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Recall.

How many tools does the Recall MCP server expose? +

8 tools across 4 categories: Destructive, Execute, Read, Write. 4 are read-only. 4 can modify, create, or delete data.

How do I enforce a policy on Recall? +

Register the Recall MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every Recall tool call.

Deterministic rules across all 8 Recall tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON RECALL →

Instant setup, no code required.

8 Recall tools catalogued and risk-classified — across an index of 43,000+ MCP servers.