AI agents use hold to create or update resources in Ombre Brain — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Ombre Brain environment.
This tool writes/creates a new memory entry into persistent storage and may merge it with existing memory buckets. It modifies stored data but is reversible in principle (memories can be deleted or updated). No code execution, financial transaction, or irreversible destruction is implied.
From the tool's definition 存储单条记忆。自动打标+合并相似桶 (Store a single memory. Auto-tag + merge similar buckets.)
Documented attack patterns abuse exactly the kind of access hold gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Ombre Brain, and nothing reaches the server without passing your rules. This is the rule we recommend for hold:
{
"version": "1",
"default": "deny",
"tools": {
"hold": {
"limits": [
{
"counter": "hold_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}hold stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
存储单条记忆。自动打标+合并相似桶。tags 逗号分隔，importance 1-10。. It is categorised as a Write tool in the Ombre Brain MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Ombre Brain MCP server in PolicyLayer and add a rule for hold: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Ombre Brain. Nothing to install.
hold is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the hold rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for hold. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
hold is provided by the Ombre Brain MCP server (p0lar1zzz/ombre-brain). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 6 Ombre Brain tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
6 Ombre Brain tools catalogued and risk-classified — across an index of 42,500+ MCP servers.