Back to skill

Security audit

dfs-optimizer

Security checks for vulnerabilities and agentic risk

Overview

The skill performs local fantasy-sports calculations without sensitive access, but its examples produce incomplete lineups despite claiming optimal results.

Treat this as demonstration code rather than a validated optimizer. Independently check roster completeness, current platform rules, salaries, and projections before using results in paid contests. The optimality claims are unsupported, and FanDuel/MLB lineup builders are absent.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The code is related to fantasy lineup selection, but materially falls short of its declared optimization purpose. It neither searches for an optimal valid lineup nor validates that all roster slots are filled, and its platform and sport support is substantially narrower than advertised. No unrelated resource access or additional undeclared side effects are evident.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The code is related to the declared DFS purpose, but its central optimization claim is materially inaccurate. It fills positions sequentially without reserving salary for remaining slots, backtracking, or validating roster completeness. The provided player pool cannot form a complete nine-player lineup under $50,000, yet the script does not report infeasibility. No unrelated resource access or undeclared external side effects are evident.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

Although ownership analysis is related to daily fantasy sports and could support lineup selection, this chunk only produces a player ranking and strategy tips. It neither constructs lineups nor performs the constrained optimization central to the declared purpose. It accesses no external resources and exhibits no unrelated or malicious behavior.

Content

No source excerpt is available for this finding.

The advertised optimal-lineup builder uses greedy selection that can return incomplete, non-optimal rosters.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The manifest at L003 promises optimal lineups maximizing projected points under salary constraints, and L017 explicitly claims mathematical optimality. Instead, L057–L076 select players by points-per-dollar without backtracking or verifying that all roster slots were filled; the NFL implementation at L124–L162 has the same limitation. These implementations can exhaust the budget before completing a roster, including with the supplied samples, yet label their output an 'Optimal Lineup' at L080 and L166.

Content

No source excerpt is available for this finding.