Back to skill
Skillv0.4.0
ClawScan security
Blockchain Analysis · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 4, 2026, 9:39 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill is an instruction-only Blockscout reference that consistently documents how to query Blockscout MCP/REST endpoints and does not request unrelated credentials, installs, or privileged access.
- Guidance
- This skill is a documentation/reference wrapper for Blockscout and appears internally consistent. Things to consider before installing: 1) it will direct the agent to call Blockscout MCP (https://mcp.blockscout.com) and Chainscout (https://chains.blockscout.com) — only install if you trust those endpoints; 2) the skill itself requests no secrets, but some Blockscout endpoints accept optional apikey/key parameters — provide those only to trusted scripts; 3) because it’s instruction-only and the agent may invoke skills autonomously, you may want to restrict autonomous execution or review agent prompts that would cause network calls if you have strict data-exfiltration or network policies; 4) absence of code files reduces immediate disk-write risk, but the skill can still instruct the agent to perform network calls — treat external network access as the main potential risk.
Review Dimensions
- Purpose & Capability
- okName/description (blockchain analysis for Blockscout MCP) match the included instructions and large set of Blockscout API reference files. There are no unexpected env vars, binaries, or config paths requested that would be unrelated to on-chain data retrieval.
- Instruction Scope
- okSKILL.md instructs the agent to use Blockscout MCP native tools, the direct_api_call proxy, and Chainscout for chain resolution. It recommends calling unlock_blockchain_analysis before other MCP calls — this is consistent with the skill's stated purpose. Instructions do not request reading local files, unrelated environment variables, or exfiltrating data to unexpected endpoints; external network usage is limited to Blockscout/Chainscout endpoints, which is expected.
- Install Mechanism
- okNo install spec and no code files that would be written to disk. Instruction-only skills present the lowest install risk and this skill follows that pattern.
- Credentials
- okThe skill declares no required environment variables, credentials, or config paths. The references do mention optional API keys/keys for Blockscout endpoints (typical for rate limiting/restricted endpoints) but the skill itself does not demand any secrets from the agent environment.
- Persistence & Privilege
- okalways is false and model invocation is allowed (default). The skill does not request persistent system-wide privileges or modify other skills. Autonomous invocation is normal and the skill's scope is limited to guidance/reference for Blockscout calls.