txcloud-diagnostics

用于腾讯云云产品异常诊断。当用户反馈腾讯云相关的任何异常、产品/实例不可用等情况时，根据反馈的实例和异常信息，自动拉取监控等数据进行分析诊断，输出原因和建议。

Key things to consider before installing: - The package did not declare required binaries or credentials, but the scripts require the tccli CLI and valid Tencent Cloud credentials (used to call monitor, tat, and other APIs). Make sure tccli is installed and be aware this skill will prompt you to authenticate. - The auth flow starts a background daemon that writes/reads files under /tmp and asks the user to paste the interactive auth code; that code will be written verbatim to /tmp and submitted to tccli. Only proceed if you trust the skill and are comfortable running such a flow locally. - The skill can perform TAT remote execution on your instances. Although the documentation lists a read-only command whitelist, the code will accept and forward arbitrary commands provided at runtime — this is a powerful capability. Only run this skill with credentials scoped to read-only/diagnostic operations, or require manual review before executing any TAT command. - The SKILL.md instructs the agent to suppress intermediate errors and internal tccli output; this can hide failures or unexpected behavior. Prefer to run the scripts manually first to observe outputs and logs (/tmp/tccli_daemon.log, /tmp/tccli_auth_link.txt) before enabling automated/agent invocation. Recommended actions: 1. Do not enable autonomous invocation unless you trust the skill fully. Limit to user-invoked only. 2. Inspect and run the scripts in a safe/test account or sandbox before using on production credentials or instances. 3. Restrict credentials used by this skill (least privilege): monitoring + describe permissions, and avoid granting broad TAT/instance-control permissions unless necessary. 4. Ask the author to update metadata to declare tccli as a required binary and to document required IAM scopes; request code changes to enforce a command whitelist for TAT at runtime rather than relying solely on documentation. If you want, I can produce a concise checklist or suggested minimal IAM policy that would be appropriate for safely running this diagnostic skill.

Type: OpenClaw Skill Name: txcloud-diagnostics Version: 1.0.0 The skill bundle provides Tencent Cloud diagnostic capabilities but includes high-risk features, most notably the ability to execute remote shell commands on user cloud instances via the Tencent Automation Tool (TAT) as seen in `prefetch.py` and `compute.md`. The instructions in `SKILL.md` and `compute.md` explicitly direct the AI agent to autonomously trigger these remote commands based on its own analysis of monitoring data, creating a significant risk if the agent is targeted by prompt injection. Furthermore, `scripts/tccli_auth_daemon.py` uses pseudo-terminals and insecure temporary files in `/tmp` to handle authentication, which is a complex and potentially fragile mechanism for managing cloud credentials.