Merge multiple observations into one. Atomic operation: provide merged text + list of observation IDs. Creates the merged observation, deletes the originals, handles embeddings. Use after consolidate to act on clusters.
AI agents use merge to create or update resources in Hippocampus — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Hippocampus environment.
While the tool performs deletion ('deletes the originals'), this deletion is part of an atomic merge operation and not the primary destructive purpose — the main intent is to consolidate data reversibly. The tool creates a new record and removes redundant predecessors.
From the tool's definition Tool description states it 'Creates the merged observation, deletes the originals' — this involves both creation (write operation) and deletion of data records.
Documented attack patterns abuse exactly the kind of access merge gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Hippocampus, and nothing reaches the server without passing your rules. This is the rule we recommend for merge:
{
"version": "1",
"default": "deny",
"tools": {
"merge": {
"limits": [
{
"counter": "merge_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}merge 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.
Merge multiple observations into one. Atomic operation: provide merged text + list of observation IDs. Creates the merged observation, deletes the originals, handles embeddings. Use after consolidate to act on clusters. It is categorised as a Write tool in the Hippocampus MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Hippocampus MCP server in PolicyLayer and add a rule for merge: 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 Hippocampus. Nothing to install.
merge 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 merge 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 merge. 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.
merge is provided by the Hippocampus MCP server (karrolcia/hippocampus). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Hippocampus, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
11 Hippocampus tools catalogued and risk-classified — across an index of 43,000+ MCP servers.