Back to skill
Skillv1.0.0
ClawScan security
Sovereign Codebase Onboarding · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignFeb 23, 2026, 11:18 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's requested capabilities, runtime instructions, and lack of install/credential requirements are consistent with a codebase-onboarding assistant.
- Guidance
- This skill appears coherent for repository analysis: it reads manifests, source files, CI, and Docker/README to build an onboarding guide and asks for no credentials or installs. Before installing, consider: (1) only grant the agent access to repositories you are comfortable having analyzed (don’t expose private secrets like .env files or credentials), (2) review the full SKILL.md/EXAMPLES to confirm it won't be pointed to any external endpoints in your deployment, and (3) if your agent platform allows scoping skills to specific repos or read-only file sets, restrict access to limit exposure. If you want extra assurance, test the skill on a public or sanitized repo first.
Review Dimensions
- Purpose & Capability
- okName/description (codebase onboarding, architecture mapping, guides) match the SKILL.md phases and the examples. The skill only describes reading repository manifests, CI, Dockerfiles, directory trees, and code to build an onboarding guide—these requirements are proportional to its purpose.
- Instruction Scope
- noteSKILL.md instructs the agent to inspect repository files (manifests, CI, Docker, entry points, configs) which is expected. One minor narrative phrase in the Philosophy section references reading "memory files" and "the journal" (personal anecdote) — this is ambiguous but the actionable instructions that follow are limited to repository artifacts. No instructions request unrelated system-wide files, credentials, or outbound exfiltration.
- Install Mechanism
- okInstruction-only skill with no install spec, no downloads, and no code files to execute. This minimizes surface area; nothing is written to disk or installed by the skill itself.
- Credentials
- okThe skill declares no required environment variables, no credentials, and no config paths. That is proportional for a purely analytical onboarding skill that works by reading repository contents.
- Persistence & Privilege
- okalways:false (default) and the skill does not request persistent system-wide changes or modification of other skills. Autonomous invocation is allowed by platform default but is not combined with broad privileges here.