predict-worknet

Fail

Audited by Snyk on May 6, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E005: Suspicious download URL detected in skill instructions.

  • Suspicious download URL detected (high risk: 0.80). Suspicious — both links point to code on unverified GitHub accounts (a raw install.sh intended to be piped to sh and a wallet repo to clone/npm-install), which can execute arbitrary code and should be treated as high risk unless you verify the authors, inspect the scripts, and validate signatures/checksums in a safe environment.

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

  • Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to run predict-agent (see SKILL.md) which fetches market context/klines, orderbook data, _internal.next_command and math "challenge" text from remote coordinator/AWP APIs (e.g., PREDICT_SERVER_URL https://api.agentpredict.work and https://api.awp.sh), and those untrusted third‑party responses are read and must be followed or used to form submissions, so external content can directly influence tool use and next actions.

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


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

  • Direct money access detected (high risk: 1.00). The skill is explicitly designed for crypto market operations: it directs the agent to use the predict-agent CLI to fetch market context, submit predictions (orders) with tickets, size positions, cancel orders, check orders and wallet status, and even create/manage an awp-wallet. Those are concrete, purpose-built commands for placing and managing market bets and interacting with a wallet/coordinator (i.e., crypto/market-order functionality), not generic tooling. This grants direct financial execution capability.

Issues (4)

E005
CRITICAL

Suspicious download URL detected in skill instructions.

W011
MEDIUM

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

W012
MEDIUM

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

W009
MEDIUM

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

Audit Metadata
Risk Level
CRITICAL
Analyzed
May 6, 2026, 02:03 AM
Issues
4