Tencent Cloud RCE Risk Control

Problem

RCE risk operations in Tencent Cloud often require many manual console or API steps: checking request overviews, analyzing risk trends, tracing a ReqId, maintaining name lists, and adjusting policies or monitoring rules. If every operation requires reading docs, assembling parameters, handling authentication and error codes, then investigating false positives, producing reports, or importing lists becomes fragmented. This skill turns those repetitive actions into natural-language commands, such as asking for the request overview for today's registration risk or importing several phone numbers into a blocklist.

How It Works

It sits on top of Tencent Cloud RCE cloud APIs and translates user intent into permission-constrained API 3.0 calls, then returns structured results. Core capabilities include:
- Risk posture: request overviews, risk trends, policy effect summaries, and hit analysis by rule, policy, or event.
- Risk lookup: query user risk profiles by AccountType and AccountId, or trace a single request by ReqId.
- Policies and lists: policy CRUD, cloning, batch creation by service; create, import, query, and expire blocklists or allowlists.
- Monitoring and data: monitoring rule CRUD, risk monitoring management, trigger chain lookup, report tasks, and decrypted content queries.
Authentication uses amccli in OAuth mode and does not ask users for long-term keys. Actual permissions are bounded by the authorized Tencent Cloud sub-account's CAM policies. Execution must still follow the internal execution manual for parameter shaping, string input rules, BusinessSecurityData wrapping, and other SOPs, rather than calling APIs from the high-level summary alone.

Scope and Limits

The skill is useful for risk operations, policy engineers, business developers, and risk product managers who perform daily inspection, policy tuning, case analysis, and trend reporting. It does not replace the RCE console and is not a general-purpose risk-control solution. Sensitive-field interfaces should be used under existing compliance policies, and batch imports or task-based operations should be verified by task IDs and failure details.