Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Freelance Automation Gig
v1.0.0自動幫客戶完成 Upwork/Fiverr 任務 - RPA、Data Entry、Web Scraping、PDF 生成等
⭐ 0· 805·10 current·12 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Stated purpose (search/apply for Upwork/Fiverr jobs and execute automation tasks) is plausible for tools like curl/node/python which the skill lists. However, the skill claims it can post proposals, check payments, and otherwise interact with platform accounts but declares no credentials, APIs, or config paths. The example commands ('upwork', 'fiverr') are not standard system binaries and appear to be placeholders, creating a mismatch between claimed capability and the concrete runtime requirements.
Instruction Scope
SKILL.md gives broad, open-ended runtime instructions (search jobs, generate proposals, execute tasks) and explicitly instructs the agent to conceal that it is an AI ("絕對唔好透露自己係 AI（扮 human freelancer）"). That directive is ethically problematic. The instructions are vague about what ’execute task’ entails, where data is sent, and how platform interactions/authentication are handled. There are no limitations or explicit safety boundaries for scraping, account actions, or handling client payment issues.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest installation risk. It simply lists required binaries (curl, node, python), which are reasonable for the described automation tasks.
Credentials
The skill requests no environment variables or credentials, yet the described tasks (posting proposals, checking payment status, operating email/Telegram/Discord bots) typically require sensitive credentials or API tokens. The absence of declared credential inputs is inconsistent and suggests either incomplete metadata or a design that would prompt the agent to request or scrape secrets at runtime — both concerning.
Persistence & Privilege
The skill does not request persistent/always-on presence (always: false) and has no install-time modifications. Autonomous invocation by the agent is allowed by platform default (disable-model-invocation: false). Combined with the instruction to impersonate humans, autonomous operation could increase risk, but the skill itself does not request elevated persistence privileges.
Scan Findings in Context
[no_code_files_analyzed] expected: The skill is instruction-only (SKILL.md) so the regex scanner had no files to analyze. Absence of findings is expected but provides limited assurance — the SKILL.md content is the primary surface to evaluate.
What to consider before installing
This skill raises multiple red flags you should resolve before installing. Ask the publisher to clarify exactly how platform interactions work: which credentials or API keys are required, how they are provided/stored, and whether any network endpoints receive client data. Do not share Upwork/Fiverr passwords or tokens unless there is a clear, auditable reason and secure storage (and be aware of platform Terms of Service — automating proposals or impersonating humans may violate them). Insist the skill remove or justify the instruction to hide that it's an AI; that is an ethical/legal risk. If you still evaluate it, test in a sandbox account, require manual approval for posting proposals and payments, and limit any tokens to least-privilege scopes. If the author cannot explain where credentials are kept, what endpoints are contacted, or why no env vars are declared despite needing platform access, treat the skill as unsafe to use.Like a lobster shell, security has layers — review code before you run it.
latestvk9706925bphyq7zwx476nnrap181ecz4
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binscurl, node, python