Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Corlious
v1.0.0Help users cook with recipes, technique explanations, and ingredient substitutions.
⭐ 0· 136·0 current·0 all-time
by@carl32
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The SKILL.md provides cooking and recipe guidance which matches the stated description, and the skill requests no binaries, installs, or credentials — that is proportionate. However, there are metadata mismatches: the registry entry labels the skill 'Corlious' (ownerId kn79...), while SKILL.md identifies as 'Chef' and _meta.json lists a different ownerId/slug (kn73..., slug 'chef'). These inconsistencies suggest the package may have been renamed, republished, or tampered with.
Instruction Scope
The runtime instructions are limited to asking users questions and providing recipe/technique guidance. They do not instruct reading files, accessing environment variables, calling external endpoints, or performing other system actions.
Install Mechanism
No install specification or code files are present — this is instruction-only and does not write code to disk or fetch external packages, which minimizes install-time risk.
Credentials
The skill declares no required environment variables, credentials, or config paths. There are no apparent requests for secrets or unrelated permissions.
Persistence & Privilege
The skill does not request always:true or other elevated persistence. It is user-invocable and allows model invocation (the platform default), which is expected for a functional skill.
Scan Findings in Context
[unicode-control-chars] unexpected: Control/unicode-control characters in SKILL.md are unexpected for a recipe guide and can be used for prompt-injection or to obfuscate content. The SKILL.md text itself appears normal, but the presence of these characters is a signal to inspect the raw file for hidden directives or tampering.
What to consider before installing
This skill's instructions are benign and require no credentials or installs, but exercise caution because the package metadata does not line up with the files inside and a prompt-injection pattern was detected. Before installing or enabling: 1) verify the publisher/owner in the registry matches the _meta.json and SKILL.md (ask the registry admin or publisher for provenance). 2) Inspect the raw SKILL.md for hidden/control characters or unusual whitespace (the unicode-control-chars finding). 3) Only enable the skill if you trust the source; because it's instruction-only the immediate technical risk is low, but metadata mismatch could indicate repackaging or tampering. 4) If you proceed, run it in a low-privilege/sandboxed environment and avoid providing any sensitive data during testing. If you cannot confirm origin, prefer not to install.Like a lobster shell, security has layers — review code before you run it.
latestvk97c0kmpkc6849hpwfageqcnc18379aw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
👨🍳 Clawdis
OSLinux · macOS · Windows