Skill flagged — suspicious patterns detected

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

officecli-academic-paper

v1.0.1

Use this skill when the user wants to create an academic paper, research paper, white paper, technical report, policy brief, or any formally structured docum...

0· 58·0 current·0 all-time
by瓦砾@iceyliu
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
high confidence
Purpose & Capability
The skill's name, description, and detailed instructions all align with building formally structured .docx documents and therefore legitimately require an officecli-like binary. However, the SKILL.md assumes availability of a third-party binary and embeds an imperative install/upgrade procedure (curl + bash) that is not declared in the registry install metadata — a mild inconsistency between claimed packaging and runtime requirements.
!
Instruction Scope
The runtime instructions explicitly instruct the agent to fetch and execute remote scripts and to call network endpoints (curl to raw.githubusercontent.com and to the GitHub API). They require the agent to run arbitrary shell commands (install/update scripts) and to perform auto-upgrades before each use. While the document-manipulation commands themselves stay on-purpose, the unconditional download-and-execute step grants the agent broad ability to run arbitrary code fetched at runtime and to modify the host environment.
!
Install Mechanism
There is no declared install spec, but SKILL.md tells the agent to run curl -fsSL https://raw.githubusercontent.com/iOfficeAI/OfficeCli/main/install.sh -o /tmp/officecli_install.sh && bash /tmp/officecli_install.sh. Downloading and executing an arbitrary install script at raw.githubusercontent.com (even from a GitHub repo) is high risk because the script can run any commands. The SKILL.md also instructs automated upgrades using the same mechanism.
Credentials
The skill does not request any environment variables, credentials, or config paths. The lack of secret/credential requests is appropriate for a document-creation tool and reduces direct exfiltration risk. However, the install/update commands could still cause indirect risk by executing code with the agent's privileges.
!
Persistence & Privilege
The skill does not set always:true, but the instructions will install or upgrade a system binary at runtime and potentially make persistent changes on the host (files under /usr/local, /tmp execution, etc.). If the agent is allowed to invoke the skill autonomously, it could cause repeated network fetch-and-execute cycles. This persistent modification behavior is not surfaced in the registry metadata.
What to consider before installing
This skill appears to be what it says (a recipe for producing academic .docx files), but the SKILL.md tells the agent to download and execute an install script from raw.githubusercontent.com and to auto-upgrade from GitHub before each use. That is the main risk. Before installing or running this skill: - Do not run the curl | bash command blindly. Inspect the install.sh contents at https://raw.githubusercontent.com/iOfficeAI/OfficeCli/main/install.sh to confirm it is safe and from a trusted maintainer. - Prefer installing officecli manually from a vetted package or repository (or pin a specific, signed release) rather than allowing automatic unpinned installs/upgrades. - If possible, run the skill in an isolated environment (container or VM) so the remote install cannot affect your primary system. - If you do not want the agent to fetch or execute remote code automatically, remove or comment out the install/upgrade block in SKILL.md or configure the agent to require explicit confirmation before running shell commands that fetch remote scripts. - Consider asking the skill author to provide an explicit install spec (packaged release URL or package-manager entry) and to avoid curl|bash auto-updates in runtime instructions. Given these factors, treat this skill as suspicious until you verify the install script and are comfortable with its runtime network/execution behavior.

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

latestvk9704nj3t0jf3bbv282vsd4f39840m4f

License

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

Comments