clanker-plugin

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill fetches and parses data from public third-party endpoints (notably https://clanker.world/api in src/api.rs and the RPC/explorer URLs listed in plugin.yaml), and the SKILL.md explicitly instructs the agent to parse quickstart JSON fields like next_command/onboarding_steps and act on them, so untrusted external content can directly influence subsequent tool use and decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill's pre-flight installation fetches and executes remote installer scripts and binaries at runtime (e.g. https://raw.githubusercontent.com/okx/onchainos-skills/${LATEST_TAG}/install.sh and https://github.com/okx/plugin-store/releases/download/plugins/clanker-plugin@0.2.5/clanker-plugin-...), which downloads and runs remote code that the skill requires to operate.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill is explicitly a blockchain financial tool: it performs on-chain write operations (deploy-token calls deployToken on the Clanker V4 factory via onchainos wallet contract-call; claim-rewards calls collectFees on the ClankerFeeLocker via onchainos wallet contract-call). It requires wallet login, ETH for gas, and reports tx_hash/explorer URLs. These are direct crypto/blockchain transaction and wallet-signing capabilities (deploying contracts and claiming on-chain rewards), so it grants direct financial execution authority.