Skill flagged — suspicious patterns detected

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

TencentCloud OPS

v1.2.0

提供腾讯云 CVM 云服务器和 COS 对象存储的创建、管理、成本控制及安全配置功能。

0· 87·1 current·1 all-time
bysuperStupidBear@ugpoor
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 skill's name, description, SKILL.md and code consistently implement CVM and COS management (creating instances, buckets, etc.), which legitimately requires Tencent Cloud API credentials. However, the registry metadata lists no required environment variables or primary credential while both the SKILL.md and src/tencent_ops.py require TENCENT_SECRET_ID and TENCENT_SECRET_KEY. That mismatch between declared metadata and actual requirements is an incoherence.
!
Instruction Scope
SKILL.md instructs the agent/user to install the Tencent CLI, create subusers, create broad policies (e.g., action lists including "cvm:*", "vpc:*", "cbs:*", and "name/cos:*" with resource "*") and to place secrets in config/.env, then run python3 src/verify_config.py. The code file uses environment variables and SDKs consistent with these actions. Problems: the instructions reference files/scripts that are not present in the manifest (config/.env.example and src/verify_config.py are mentioned but not included), and the policy examples use wildcard resources and broad action scopes which increase risk if applied to non-isolated accounts.
Install Mechanism
There is no install spec in the registry (instruction-only), and SKILL.md recommends installing standard Python packages (tencentcloud-sdk-python, cos-python-sdk-v5, python-dotenv) and optionally tccli via Homebrew. These are expected for Tencent Cloud SDK usage and are proportionate. No unknown download URLs or archive extraction are used.
!
Credentials
The skill needs cloud API credentials (TENCENT_SECRET_ID and TENCENT_SECRET_KEY) and region settings according to SKILL.md and the code. That is proportionate for cloud resource management, but the registry metadata fails to declare them. The provided policy examples grant broad privileges and use resource "*" which is wider than least-privilege best practice; the guidance recommends using a subuser, but the examples still create very permissive policies. Also some names (e.g., 'name/cos:*') look unusual and should be verified against Tencent's policy action syntax.
Persistence & Privilege
The skill does not request always: true and is user-invocable with normal model invocation allowed (default). It does not include an install spec that writes arbitrary binaries or attempts to modify other skills' configuration. Note that, like any cloud-management skill, if granted credentials the agent (when invoked) can perform destructive actions — this is normal for such a tool but increases blast radius if credentials are over-privileged.
What to consider before installing
Things to check before installing or running this skill: - Expectation mismatch: SKILL.md and src/tencent_ops.py require TENCENT_SECRET_ID and TENCENT_SECRET_KEY, but the registry metadata did not declare any required credentials. Treat that omission as a red flag—the skill will need API keys to function. - Use least privilege and isolation: create a dedicated sub-user in a test/isolated Tencent Cloud account and grant only the minimal actions for the specific resources you want to manage. Avoid applying the provided example policies with action lists and resource "*" in a production account. - Verify missing files: SKILL.md references config/.env.example and src/verify_config.py, but those files are not present in the provided manifest. Confirm the repository includes those files and inspect them before running any verification scripts. - Audit the code: although the included src/tencent_ops.py appears to use official Tencent SDKs and reads env vars directly (no obvious obfuscation or external exfil endpoints in the truncated file), review the full source (including truncated sections) to ensure there are no hidden network callbacks, logging of secrets, or unexpected endpoints. - Do not use root/main account keys: follow the skill's own advice to use a sub-user and do not store long-lived keys in public repos. Add config/.env to .gitignore and rotate keys regularly. - Confirm policy action syntax: some policy action strings in the SKILL.md (e.g., "name/cos:*") look unusual — verify they match Tencent CAM action names before applying. If you are not comfortable auditing the code yourself, run this tool only in a controlled, non-production account and with keys that have tightly scoped permissions.

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

cloudvk979ngyjfe7gctz7pzw4x5r7js83tdrfdevopsvk979ngyjfe7gctz7pzw4x5r7js83tdrflatestvk97e969eyz0jr2e8sardk7zz1x83vnfhopsvk979ngyjfe7gctz7pzw4x5r7js83tdrftencentcloudvk979ngyjfe7gctz7pzw4x5r7js83tdrf

License

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

Comments