Privileged always
Warn
- Finding
- Skill is configured with always=true (persistent invocation).
AI Control Protocol (Anti-Sycophancy & Zero-BS)
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This instruction-only skill persistently changes the agent’s reasoning and response style, but it does not install code, request credentials, access files, or call external services.
This skill is not showing evidence of malicious behavior, but it is designed to be always on and to make the assistant more confrontational, skeptical, and structured. Install it if you want that persistent style; avoid it if you prefer neutral, concise, or user-led responses.
Static analysis
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
What this means
The assistant may become more critical, less affirming, and less direct in planning or advice tasks.
Why it was flagged
The skill intentionally imposes global response rules that can redirect normal assistant behavior, such as adding uncertainty labels, challenging premises, and delaying strategic recommendations.
Skill content
These rules must be executed in every single conversational or analytical output.
Recommendation
Install only if you want persistent anti-sycophancy behavior across conversations; disable it if it interferes with direct-answer workflows.
What this means
The skill can affect every compatible interaction after installation.
Why it was flagged
The skill is persistently invoked by design. The artifact discloses this and provides a purpose-based justification, and there is no evidence of hidden background code or autonomous external activity.
Skill content
always: true
Recommendation
Review the behavioral rules before installing and keep the skill enabled only if you want this always-on reasoning style.