Back to skill
Skillv1.0.0
ClawScan security
Drunk Mode · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 30, 2026, 1:20 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This is an instruction-only persona toggle that reads/writes a local state file to enable a ‘drunk’ reply style; it doesn't request credentials, network access, or install code and is internally consistent aside from a minor metadata omission.
- Guidance
- This skill is low-risk: it only reads/writes a small JSON file in the skill workspace and changes reply style when enabled. Before installing, consider: (1) confirm you trust the skill source (the package has no homepage and an unknown owner), (2) be aware it will create/update ~/.openclaw/workspace/skills/drunk-mode/drunk_state.json to store enabled/level, and (3) if you prefer, inspect or sandbox the skill directory or remove the file after use. No credentials or network access are requested. If you want stronger assurance, ask the publisher for provenance or source code before enabling.
Review Dimensions
- Purpose & Capability
- noteThe skill's name/description (a togglable 'drunk' persona) matches what the SKILL.md instructs: reading/writing a local state file and altering reply style. Minor inconsistency: the SKILL.md references a concrete config path (~/.openclaw/workspace/skills/drunk-mode/drunk_state.json) but the registry metadata lists no required config paths. This is likely harmless (the skill only uses its own workspace file) but is an undeclared file-access detail the user should be aware of.
- Instruction Scope
- okInstructions are narrowly scoped: they tell the agent to set/read the single state JSON file and to apply the drunk persona to replies when enabled. There are no instructions to read other files, access environment variables, or send data to external endpoints. The persona override applies to replies when enabled — that is expected behavior for a persona skill.
- Install Mechanism
- okNo install spec and no code files (instruction-only). Nothing is downloaded or written beyond the state file in the skill workspace included in the manifest.
- Credentials
- okThe skill requests no environment variables, keys, or credentials. Its only persistent artifact is a simple JSON state file (provided in the bundle) used to track enabled/level.
- Persistence & Privilege
- okalways is false and the skill doesn't request elevated privileges or to modify other skills or system-wide settings. It will read/write its own state file in the skill workspace when toggled, which is normal for a persona toggle.