Skillv98.0.1

ClawScan security

Agent Prompt Patterns · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 17, 2026, 5:22 AM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: This is an instruction-only collection of production prompt and agent-operating patterns that aligns with its description and does not request credentials, install code, or ask for unrelated system access.
Guidance: This skill is a handbook of production agent patterns and is internally consistent, but review before enabling it for an agent that has filesystem or tool access: the guidance explicitly encourages gathering 'proof' (file paths, command outputs, PIDs, etc.), which could surface sensitive data if your agent can read files or execute commands. Only enable these patterns for agents running in environments where you trust and limit their access (sandboxed containers, least-privilege tool permissions). Also confirm you trust the source (no homepage or maintainer link provided) and consider copying the guidance into your own vetted AGENTS.md rather than giving the skill autonomous high-privilege runtime access.

Purpose & Capability: okName/description match the SKILL.md content: a set of production prompt patterns, operational checklists, and example snippets. It does not declare unrelated binaries, env vars, or config paths.
Instruction Scope: noteThe instructions recommend collecting 'proof' (PIDs, file paths, command output, URLs) and include shell-snippet examples. That is coherent for production agent patterns but implies the agent *should* have access to the host filesystem/commands to implement the pattern; the skill itself does not request or provision that access. If you grant the agent tool/file access, expect it to attempt these kinds of reads/commands.
Install Mechanism: okNo install spec and no code files — lowest-risk delivery (instruction-only). Nothing is downloaded or written to disk by an installer in the package metadata.
Credentials: okNo required environment variables, credentials, or config paths are declared. Patterns mention typical artifacts (file paths, process IDs) for proofs of work but do not request keys or unrelated secrets.
Persistence & Privilege: okalways is false and model invocation is allowed (platform default). The skill does not request elevated persistence or modify other skills' configs.