Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Tencent Cloud TIONE

v1.0.5

腾讯云 TI-ONE 训推平台查询工具集,支持训练任务、在线服务、开发机、资源组、模型仓库、数据集、日志、事件等模块的查询操作。

0· 222·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The scripts and SKILL.md implement a read-only TI-ONE query tool that calls tccli Describe* APIs and requires Tencent Cloud API keys and a default region. However the registry metadata earlier in the listing claimed no required env vars / primary credential while SKILL.md declares TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY as required — this mismatch is unexplained and inconsistent with the stated provisioning metadata.
Instruction Scope
SKILL.md explicitly constrains runtime behavior to invoking provided scripts under scripts/*.sh, and the scripts themselves only call tccli Describe* APIs and format output. They read expected environment variables (credentials and region). That scope aligns with the stated read-only purpose, but the restriction ('must not call tccli directly') is advisory only — not enforced by platform — so an agent or user with execution capability could bypass it.
Install Mechanism
There is no registry install spec, but SKILL.md contains an 'openclaw' install hint recommending pip install tccli and apt/brew install jq. Recommending tccli and jq is expected for this task, but the discrepancy between declared install spec (none) and SKILL.md metadata is an inconsistency to clarify. The recommended installers (pip/apt/brew) are standard and not suspicious by themselves.
!
Credentials
The scripts require TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY and a region variable (TENCENT_TIONE_DEFAULT_REGION). Those credentials are necessary for tccli calls and are proportionate to the function. However the registry-level metadata omitted these required secrets, which is inconsistent; that could lead to unexpected behavior where the skill runs without prompting for credentials or the platform does not surface the credential requirement to users. Ensure the skill only receives least-privilege/read-only credentials.
Persistence & Privilege
The skill is not marked always:true and does not request persistent system-wide privileges. It is user-invocable and allows model invocation (default), which is expected. The skill does not modify other skills or system configs.
What to consider before installing
What to check before installing/use: - Metadata mismatch: The registry entry does not list required Tencent credentials but SKILL.md and the scripts do require TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY. Do not assume the platform will prompt for these — verify and set them yourself if you intend to use the skill. - Credentials: Provide least-privilege/read-only API keys for TI-ONE. Although scripts call only Describe* APIs, full-permission keys could be abused if an attacker or a misconfigured agent runs tccli directly. - Install: The skill expects tccli and jq. SKILL.md suggests pip/apt/brew installs; confirm you install from official sources (pip index / distro packages) and inspect the tccli package origin if you are concerned. - Enforcement: The SKILL.md claims 'do not call tccli directly' — this is a procedural constraint, not technically enforced. If you allow autonomous skill invocation, the agent could potentially run other commands unless your runtime sandbox prevents it. Consider disabling autonomous invocation for untrusted skills or review invocation policies. - Audit: Because this skill includes shell scripts, review the scripts (they are included) to confirm there is no code path that prints or transmits secret values. Verify the scripts' behavior in a safe environment with test credentials first. If you want me to, I can produce a short checklist and commands to validate the scripts locally (e.g., run lint/grep for suspicious network endpoints, simulate with dummy env vars, or verify which tccli actions are invoked).

Like a lobster shell, security has layers — review code before you run it.

latestvk972k68c71wpgdgy7s6387sqa584r979

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

ConfigTENCENT_TIONE_DEFAULT_REGION

Comments