Dangerous tool approvals via Zalo
Dangerous tool approvals via Zalo Teddy MCP does not auto-approve dangerous tools when TEDDY_AUTO_APPROVE_DANGEROUS=false. Flow Agent calls a dangerous MCP tool (e.g. teddy-live__teddy_comment_set_status). Response includes preview: true, requires_approval, approval_token, and a summary.…
Dangerous tool approvals via Zalo#
Teddy MCP does not auto-approve dangerous tools when TEDDY_AUTO_APPROVE_DANGEROUS=false.
Flow#
- Agent calls a dangerous MCP tool (e.g.
teddy-live__teddy_comment_set_status). - Response includes
preview: true,requires_approval,approval_token, and a summary. - Agent messages Zalo (zalouser) with a fixed format:
⚠️ Teddy needs approval
Tool: comment.set_status
Summary: …
Token: <approval_token>
Reply: approve <token>
- Human replies
approve <token>(or paste the token after “approve”). - Agent re-invokes the same tool with the same input plus:
approved: trueapproval_token:<token>
Tokens expire (~1 hour) and are one-time use.
Do not#
- Set
TEDDY_AUTO_APPROVE_DANGEROUS=trueon live. - Reuse a token for different input.
- Skip the preview step.