SkillPager: Query-Adaptive Intra-Skill Navigation via Semantic Node Retrieval
For developers of skill-based LLM agents, this work introduces a retrieval paradigm that significantly reduces token costs while maintaining high context sufficiency, though the gains are partly due to typed semantic granularity rather than a novel retrieval algorithm.
SkillPager addresses the problem of retrieving minimal, execution-sufficient context from long procedural documents for skill-based LLM agents. It achieves 78.89% context sufficiency while reducing prompt tokens by 47.04% compared to full-document prompting, and outperforms graph-based baselines by 12.16%.
Skill-based LLM agents increasingly rely on long procedural documents, but full-document prompting wastes tokens and dilutes information critical to execution. We study this setting as intra-skill retrieval, where the goal is to select a minimal, execution-sufficient context from a known skill document given a query. We present SkillPager, a two-stage framework that parses each Markdown skill into typed semantic nodes offline and leverages Maximal Marginal Relevance (MMR) to perform global, query-conditioned node selection online. On a benchmark of 395 skills and 1,975 queries, SkillPager achieves 78.89% LLM-judged context sufficiency, compared to 82.23% for the exhaustive full-document baseline, while reducing prompt tokens by 47.04%. A granularity ablation shows that applying the same retrieval algorithm to raw fixed-length chunks reaches a comparable 81.77% sufficiency but increases token cost by 28.81%, demonstrating that efficiency gains are driven by typed semantic granularity rather than the retrieval algorithm alone. Among graph-based baselines, SkillPager outperforms the strongest baseline by a margin of 12.16%. Further ablations show that supporting content is most effective when retained in the candidate pool and selected adaptively rather than removed by static heuristics. These results identify typed intra-document retrieval as a distinct access problem for skill-based agents.