Back to skill
Skillv1.0.0
ClawScan security
Micro SaaS Launch Pack · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 25, 2026, 1:00 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This is an instruction-only
- Guidance
- This skill is low-risk: it only contains written instructions for generating business-launch outputs and does not install code or ask for credentials. Before using, avoid pasting proprietary secrets or private customer data into prompts (the skill will echo inputs into plans). Treat generated pricing/legal/financial recommendations as starting points — verify with domain experts. If you plan to let the agent act autonomously with many skills enabled, remember that autonomous invocation is platform-default; this skill itself does not escalate privileges or access system resources.
Review Dimensions
- Purpose & Capability
- okThe name/description (Micro SaaS Launch Pack) matches the SKILL.md: the instructions only ask the agent to analyze ideas and produce launch artifacts. There are no unrelated requirements (no cloud credentials, binaries, or config paths).
- Instruction Scope
- okSKILL.md contains only content-generation steps (scoring, MVP design, landing copy, distribution plan, 7-day plan). It does not instruct the agent to read files, call external endpoints, access environment variables, or exfiltrate data.
- Install Mechanism
- okNo install spec and no code files — the skill is instruction-only, which minimizes on-disk changes and execution risk.
- Credentials
- okThe skill declares no required environment variables, credentials, or config paths and the instructions do not reference any secrets. Requested access is proportional (none).
- Persistence & Privilege
- okFlags are default (always:false, user-invocable:true, model invocation allowed). The skill does not request persistent privileges or to modify other skills or system settings.