# delete_observations

Delete specific observations from entities

Agent View of the PolicyLayer registry record for `delete_observations`. HTML page: https://policylayer.com/tools/delorenj-mcp-qdrant-memory/delete-observations

## Facts

- Tool: `delete_observations`
- Server: MCP Memory Server with Qdrant Persistence (`delorenj/mcp-qdrant-memory`) — https://policylayer.com/tools/delorenj-mcp-qdrant-memory.md
- Homepage: https://github.com/delorenj/mcp-qdrant-memory
- Risk category: Destructive (Critical risk)
- Registry record: grade F, identity unverified
- Server rate-limited: no
- Parameters: 0
- Recommended policy verdict: Hidden

## Example call (MCP tools/call, JSON-RPC 2.0)

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "delete_observations",
    "arguments": {}
  }
}
```

## Why delete_observations is rated Critical

The tool permanently removes observation data from the knowledge graph. Deletion operations are irreversible and cannot be undone, placing this in the Destructive category. While the blast radius is contained to observations (not the entire graph), unauthorized deletion of knowledge graph data is a high-severity risk for data integrity and system reliability.

From the tool's own definition: "Tool name 'delete_observations' and description 'Delete specific observations from entities' explicitly perform deletion, which is irreversible data removal."

## Use case

AI agents call delete_observations to permanently remove resources in MCP Memory Server with Qdrant Persistence, typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.

## Recommended policy (PolicyLayer)

Verdict: **Hidden**. Enforced by the PolicyLayer MCP gateway (https://policylayer.com/mcp-gateway) before a call reaches MCP Memory Server with Qdrant Persistence:

```json
{
  "version": "1",
  "default": "deny",
  "hide": [
    "delete_observations"
  ]
}
```

## Other tools on MCP Memory Server with Qdrant Persistence (7)

- `delete_entities` — Destructive — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/delete-entities.md
- `delete_relations` — Destructive — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/delete-relations.md
- `read_graph` — Read — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/read-graph.md
- `search_similar` — Read — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/search-similar.md
- `add_observations` — Write — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/add-observations.md
- `create_entities` — Write — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/create-entities.md
- `create_relations` — Write — https://policylayer.com/tools/delorenj-mcp-qdrant-memory/create-relations.md

## For agents

This record is a snapshot. Live verdicts and the full registry:

- Check every server in your MCP config at once: `npx -y policylayer stack`
- Vet a server before you add it: install the mcp-precheck skill — `npx skills add https://policylayer.com` (skill text: https://policylayer.com/skill.md)
- Query the registry over MCP: endpoint `https://api.policylayer.com/mcp` — tools `check_mcp_server`, `check_mcp_stack`, `check_tool`, `search_registry`, `get_change_events`

---

Source: the PolicyLayer MCP registry — one continuously verified record per MCP server. Full record: https://policylayer.com/registry?q=delorenj-mcp-qdrant-memory · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/delorenj-mcp-qdrant-memory