Unverified50% confidenceSolutionExact time
Instructions in a prompt are not an authorization system; the operator, the resource and the permitted actions should be checked in the tool wrapper at the last moment before execution, high-impact actions should require human approval, and read and write capabilities should be separated wherever possible.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe prompt must explicitly set up a "human approval" step to prevent the assistant from executing irreversible actions such as posting paid tasks without confirmation.72% similarUnverifiedProposals should be stored in a location writable by the executor, and approvals should be stored in a path to which the executor cannot push.71% similarUnverifiedIn n8n workflows, read-only tools (querying orders, reading product catalogs, retrieving knowledge bases, etc.) can run automatically, while sensitive tools (write operations such as refunds, address changes, order cancellations, initiating payments, etc.) require waiting for manual approval.69% similarUnverifiedThe deployment checklist includes: clearly defined output goals, least privilege, read-only or draft-only access in the early stage, retention of visible records, handling mechanisms for timeout retries and duplicate operations, human approval required before sending, deleting, or modifying, setting a cost ceiling, and providing a one-click shutdown.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1057258API
curl https://kongchang.com/api/v1/knowledge/claims/1057258MCP
get_claim(id=1057258)