菜单
Enable research agents to interleave evidence-based drafting with reasoning-driven deepening, automatically expanding outlines based on discovered gaps, using trajectory pruning for efficient RL training.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/ADu2021/skillXiv --skill agentcpm-report命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Convert arXiv papers into ready-to-use agent skills using category-aware extraction. First classifies the paper into one or more of 11 research categories, then applies a specialized extraction pipeline for each category — because different types of papers produce different types of usable knowledge. A single paper can yield multiple skills if it spans categories. Use this skill whenever the user wants to turn a paper into a skill, extract practical techniques from research, build a skill library from papers, convert arXiv papers into reusable agent instructions, or batch-process multiple papers into skills. Also trigger when someone asks about extracting actionable knowledge from papers, making research practical for LLM agents, or systematically converting academic contributions into structured agent capabilities.
Establish regret bounds for behavior cloning with discretized actions combining statistical error and quantization error terms. Prove smoothness requirements for safe quantizer design, show that learning-based quantizers fail these requirements, and propose model-based augmentation to reduce error dependence from H² to H.
Dynamically allocate LoRA ranks per-layer during fine-tuning instead of using fixed uniform ranks. Learn optimal rank for each layer and subject via variational framework with discretized exponential distribution, reducing memory footprint while maintaining fidelity and text-alignment.
基于 SOC 职业分类
正在显示 SKILL.md
| name | agentcpm-report |
| title | AgentCPM-Report: Interleaving Drafting and Deepening for Deep Research |
| version | 0.0.2 |
| engine | skillxiv-v0.0.2-claude-opus-4.6 |
| license | MIT |
| url | https://arxiv.org/abs/2602.06540 |
| keywords | ["Agent Planning","Research Automation","Iterative Refinement","Report Generation","Multi-Stage Training"] |
| description | Enable research agents to interleave evidence-based drafting with reasoning-driven deepening, automatically expanding outlines based on discovered gaps, using trajectory pruning for efficient RL training. |
Existing deep research systems separate planning from writing, limiting adaptive discovery. Agents that rigidly follow pre-planned outlines miss opportunities to refine understanding as evidence emerges. The "insight ceiling" arises from static structure: agents execute plans without evaluating draft quality or identifying knowledge gaps mid-process.
WARP (Writing As Reasoning Policy) interleaves evidence-based drafting with reasoning-driven deepening. Rather than plan-then-write, the agent autonomously:
This mirrors human knowledge-transforming processes where writing reveals what you don't know.
Core cycle (Plan-Draft-Deepen-Decide):
class ResearchAgent:
def __init__(self, query):
self.outline = create_sparse_outline(query)
self.draft = ""
def evidence_based_draft(self):
for section in self.outline:
search_query = contextual_query(self.draft, section)
docs = retrieve_documents(search_query)
section_content = synthesize(docs)
self.draft += section_content
():
gaps = identify_gaps(.draft)
confidence = analyze_coverage(.draft)
gaps confidence < threshold:
section = gaps[]
.outline = expand_section(.outline, section)
:
Training procedure (3-stage):
Trajectory pruning: Identify optimal stopping points in teacher sequences—the earliest decision point where continuing doesn't improve final report quality. This accelerates training by removing redundant steps.
When to use: Deploy for research tasks requiring iterative knowledge discovery (literature reviews, technical feasibility studies, competitive analysis). Less effective for well-structured tasks with clear boundaries.
Query structuring: Start with complex, multi-part queries; agent naturally decomposes during deepening. Avoid over-specific initial queries.
Convergence signals: Monitor draft-to-outline ratio (well-formed drafts show 5–10× expansion); terminate after 2–3 deepening cycles without gap discovery.
Model scaling: Cold-start SFT works with 8B models; smaller models may require larger teacher datasets. Scaling to 70B improves deepening quality significantly.
The framework uses atomic skill decomposition to train small models on complex agentic reasoning. Trajectory pruning removes ~40% of redundant steps without quality loss, enabling efficient RL on limited data. Small (8B) models achieve competitive performance with appropriate curriculum, demonstrating that multi-stage training is more important than model scale for agentic systems.