Skip to content
Preprint

Convergent Detour Hijacking: Task-Preserving Resource Amplification in Skill-Based LLM Agents

Aug 2026 · 1 citation · 27 references
Computer Science

TL;DR

Convergent Detour Hijacking is introduced, a text-only, runtime-independent attack that couples selection manipulation, malicious skill instructions, and tool-chain resource amplification largely separately and evaluates it across multiple LLM backends and 491 held-out tasks under single-task and multi-turn conditions.

Abstract

LLM agents increasingly rely on third-party skills, using natural-language descriptions for selection and instruction bodies for planning. This progressive-disclosure design exposes two sequential control points to untrusted publishers: a static skill may steer an otherwise correct task onto an unnecessarily costly trajectory. Prior work studies selection manipulation, malicious skill instructions, and tool-chain resource amplification largely separately, leaving their end-to-end composition unclear. We introduce Convergent Detour Hijacking (CDH), a text-only, runtime-independent attack that couples these stages. Under shared semantic cover, a description establishes relevance during selection, while an aligned body reuses that rationale to fabricate plausible dependencies during planning. CDH attracts an attacker-controlled coordinator alongside legitimate skills, recruits unnecessary benign skills into a bounded detour, and then re-enters the original route to preserve task completion. We evaluate it across multiple LLM backends and 491 held-out tasks under single-task and multi-turn conditions. On DeepSeek-V4-Pro, the matched coordinator is selected in 80.02% of tasks; among coordinator-hit runs that complete tasks, token consumption and end-to-end execution time increase by 66.91% and 92.45%, respectively, while aggregate task completion remains comparable. Thus, correct outcomes do not guarantee trajectory integrity or cost safety.

View source

Similar papers

Preprint Aug 2026

ElasticBack: Stealthy Conditional Backdoor in LLM-Agent Skills via Coupled Trigger-Rule Optimization

ElasticBack is presented, an effective conditional single-skill backdoor that plants a rule R in the skill document and a benign-looking trigger T in the user query, so the malicious payload fires only when both co-occur, and motivates stronger defenses for the skill supply chain.

Hao Sui, Si-Meng Qin, Jie Liao et al. · 0 citations
Preprint Sep 2026

A Finger on the Scale: Covert Policy Steering through Agentic Skills

SkillShift is presented, a constrained black-box framework for covert policy steering without explicit target command injection or task hijacking that combines semantically plausible policy edits with hierarchical validation, failure-guided optimization, and strategy compression to preserve effectiveness, output validi...

Jia-Rui Li, Jia-Hao Chen, Chun-Yi Zhou et al. · 0 citations
#artificial intelligence Preprint Oct 2026

Chaining Skills to Hijack LLM Agents

LLM agents use skills to improve performance on specialized tasks. To complete a user request, an agent may invoke several skills in sequence, allowing information produced under one skill to guide the next. Because skills may come from open-source repositories, this handoff can also carry attacker-controlled claims in...

Tian Dong, Zi-Xuan Ma, Haodong Zhao et al. · 1 citation

Pretext: Defeating Malicious Skill Detection Frameworks for AI Agents

A white-box LLM attacker, Pretext, iteratively crafts skills that evade detection while still delivering the payload and performing the benign task: moving the payload from code into natural language leaves static analysis inert, and splitting instructions across files keeps the LLM stage below its blocking threshold.

Tobias Kaisar, Aritra Dhar · 0 citations
Preprint Aug 2026

Daydreaming: Stealing Hidden Agent Skills through Black-Box Task Interaction

Daydreaming is presented, an execution-only attack that steals a multi-file skill through black-box task interactions and produces installable skills using a median of 32 victim calls per skill even with disclosure defenses enabled.

Yu-Lin Tsai, Yu-An Lu, Ci-Yang Tsai et al. · 0 citations
Preprint Aug 2026

SkillBloat: Token Amplification Attacks via Skill Injection in LLM Coding Agents

Results show that skill ecosystems expose a practical resource-amplification attack surface that is orthogonal to existing security-oriented skill poisoning, and shows that iterative optimization provides additional benefit beyond initial attack-type selection.

Yuan-Jin Zheng, Jing-Bang Chen · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.