Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Larry Marketing
v1.0.0Automate TikTok + Instagram slideshow marketing for any app or product. Researches competitors, generates AI images, adds text overlays, posts via Upload-Pos...
⭐ 0· 177·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The stated purpose matches the code: scripts generate images, overlay text, post via Upload-Post, and read analytics/RevenueCat. However the skill metadata declares no required env vars or primary credential, while the runtime scripts and README expect an Upload-Post API key, an image-generation API key (OpenAI / Stability / Replicate) and optionally RevenueCat secret keys placed in a config.json. That discrepancy between declared requirements and actual runtime needs is an important inconsistency.
Instruction Scope
SKILL.md and references instruct the agent to do browser-based competitor research (with user permission), generate images, write/read local config and report files (e.g., hook-performance.json, analytics-snapshot.json, rc-events.json), and call external services (api.upload-post.com and RevenueCat endpoints). These actions are within the marketing scope, but the onboarding instruction to "IMMEDIATELY start a conversation" could be intrusive if the agent is activated unexpectedly. The skill does not appear to instruct collecting unrelated system secrets or scanning unrelated directories, but it does expect the agent to read/write workspace files and network-accessible APIs.
Install Mechanism
There is no install spec (instruction-only), which reduces supply-chain risk. But the SKILL.md and README require installing Node.js and native node-canvas (which may require Python, make, and a C++ toolchain). That is expected for these scripts but increases friction and the chance of build-time surprises. No downloads from obscure URLs or extracted archives were observed in the provided files.
Credentials
The registry lists no required environment variables or primary credential, yet the code and README clearly require sensitive keys in a config.json: Upload-Post API key (required), image-gen API keys (OpenAI/Stability/Replicate) if those providers are used, and optionally RevenueCat V1 secret keys for conversion tracking. Asking for RevenueCat secret (sk_...) is legitimate for conversion correlation but is sensitive; the skill should have declared these required credentials up front and documented the minimal privileges needed. The absence of declared env vars in metadata is a mismatch and a potential safety oversight.
Persistence & Privilege
The skill does not request 'always: true' and is user-invocable with normal autonomous invocation allowed. It writes and reads files under its workspace (reports, snapshots, config), which is expected for this type of automation. It does not appear to modify global agent settings or other skills. The onboarding directive to begin conversation immediately is operationally broad but not a privileged system change.
What to consider before installing
This skill appears to implement what it claims (image generation, overlays, multi-platform posting via Upload-Post, analytics, optional RevenueCat correlation), but there are important mismatches you should address before installing:
- Metadata vs reality: The registry metadata declares no required credentials, but the scripts expect an Upload-Post API key (required) and optionally OpenAI/Stability/Replicate keys and RevenueCat secret keys in a local config.json. Do not assume the platform will inject these — you'll need to provide them and they are sensitive.
- Limit secrets: Use minimal-scope or throwaway accounts where possible (a dedicated Upload-Post account, or API key scoped for only the needed endpoints). Treat RevenueCat V1 secret keys like full-production secrets; only provide them if you understand the implications.
- Review outgoing endpoints: The code calls api.upload-post.com and RevenueCat endpoints. Verify those domains and the Upload-Post service are trustworthy for your use case before giving API keys.
- Inspect the rest of the code: Several files were truncated in the package listing. Review all scripts for any additional outgoing endpoints or hidden behavior before running.
- Run in a sandbox first: Install and run the scripts in a controlled environment (separate directory, container, or VM) to see what files are created and what network calls occur. Check network logs for unexpected destinations.
- Browser/agent behavior: The skill wants to perform browser-based competitor research and to "IMMEDIATELY start a conversation" on load. Ensure your agent is configured to ask permission before browsing and that you expect the onboarding prompt.
If you need a concise checklist to proceed: (1) manually inspect all files, (2) create dedicated API keys with minimal privileges, (3) run in a sandbox, and (4) do not provide your primary RevenueCat or production keys until you confirm the behavior and trust the Upload-Post service.scripts/check-analytics.js:37
File read combined with network send (possible exfiltration).
scripts/daily-report.js:40
File read combined with network send (possible exfiltration).
scripts/generate-slides.js:38
File read combined with network send (possible exfiltration).
scripts/post-to-platforms.js:33
File read combined with network send (possible exfiltration).
About static analysis
These patterns were detected by automated regex scanning. They may be normal for skills that integrate with external APIs. Check the VirusTotal and OpenClaw results above for context-aware analysis.Like a lobster shell, security has layers — review code before you run it.
latestvk971g4m9rs6rs7b153e5na6bkh8329p0
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
