Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems
This addresses a critical security vulnerability for users of LLM-based coding agents, as it exposes how malicious third-party skills can compromise systems despite existing safeguards, representing a novel attack vector rather than an incremental improvement.
The paper tackles the problem of supply-chain poisoning attacks against LLM coding agent skill ecosystems by introducing Document-Driven Implicit Payload Execution (DDIPE), which embeds malicious logic in skill documentation to bypass defenses, achieving bypass rates of 11.6% to 33.5% across frameworks and models.
LLM-based coding agents extend their capabilities via third-party agent skills distributed through open marketplaces without mandatory security review. Unlike traditional packages, these skills are executed as operational directives with system-level privileges, so a single malicious skill can compromise the host. Prior work has not examined whether supply-chain attacks can directly hijack an agent's action space, such as file writes, shell commands, and network requests, despite existing safeguards. We introduce Document-Driven Implicit Payload Execution (DDIPE), which embeds malicious logic in code examples and configuration templates within skill documentation. Because agents reuse these examples during normal tasks, the payload executes without explicit prompts. Using an LLM-driven pipeline, we generate 1,070 adversarial skills from 81 seeds across 15 MITRE ATTACK categories. Across four frameworks and five models, DDIPE achieves 11.6% to 33.5% bypass rates, while explicit instruction attacks achieve 0% under strong defenses. Static analysis detects most cases, but 2.5% evade both detection and alignment. Responsible disclosure led to four confirmed vulnerabilities and two fixes.