Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
prompt-nubaby
v1.1.0Nubaby prompt system for prompt augmentation, routers, dictionaries, dataset captions, prompt tags, compact prompts, video/storyboard prompt shaping, and str...
⭐ 0· 53·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description (prompt augmentation, routers, dataset captions, tags, retention expander) align with the provided SKILL.md, knowledge JSON, and a small manager script that implements a 'learn'/save function. The files and modes described are consistent with a prompt-augmentation skill.
Instruction Scope
SKILL.md explicitly instructs the assistant to 'remember' and persist user-provided master prompts into knowledge/master_recipes.json and to consult that knowledge base when augmenting prompts. Persisting arbitrary user input is within the skill's stated function but expands scope to durable storage of whatever users send (including potential secrets, PII, or problematic content). The SKILL.md does not define sanitization, review, or consent workflows for stored content.
Install Mechanism
No install spec; skill is instruction-first and ships only static files and a tiny script. There are no downloads, external installers, or extracted archives. Low install risk.
Credentials
The skill requests no environment variables or external credentials (proportional). However, it requires file-system write access to knowledge/master_recipes.json (implemented by scripts/manager.py), which means any sensitive user input saved by the skill will persist locally. The skill does not ask for unrelated credentials, but persistence of arbitrary user content increases privacy risk.
Persistence & Privilege
always:false (no forced global enable) and autonomous invocation is default platform behavior. The skill explicitly promises permanent accumulation of user-provided prompts into a bundled knowledge file. That persistent storage combined with autonomous invocation increases blast radius for sensitive or disallowed content (stored prompts could later be used automatically). The script lacks input validation or access controls.
What to consider before installing
This skill appears to do what it says (prompt augmentation) and has no network calls or credential requests, but it will permanently save any prompt you tell it to "remember" into knowledge/master_recipes.json. Before installing or enabling: 1) decide whether you are comfortable having prompts (which might include PII, API keys, or sensitive text) stored on disk; 2) inspect and back up knowledge/master_recipes.json after use and delete any entries you don't want persisted; 3) consider limiting the agent's file-write capability or running the skill in a sandboxed environment; 4) the manager script has no sanitization or consent flow — avoid telling the skill to remember secrets; 5) note the content contains sexualized portrait examples (some mention ages) — verify that the skill's safety rules meet your policy needs. If you need the skill but want lower risk, ask for a version that requires explicit confirmation before writing and that strips or reviews sensitive fields before saving.Like a lobster shell, security has layers — review code before you run it.
augmentationvk97a4txcnm1gwfkdv2x11a4n2983n1bslatestvk97a4txcnm1gwfkdv2x11a4n2983n1bsnubabyvk97a4txcnm1gwfkdv2x11a4n2983n1bspromptvk97a4txcnm1gwfkdv2x11a4n2983n1bsretention-shotvk97a4txcnm1gwfkdv2x11a4n2983n1bs
License
MIT-0
Free to use, modify, and redistribute. No attribution required.