| name | paper-scout |
| description | Daily paper scout for Auto-Skill and Auto-Rubric research. Use when: searching for new papers on self-evolving agents, skill evolution, rubric learning, preference alignment, reward modeling, agentic evolution. Searches arxiv for latest papers, recommends noteworthy ones, and updates the Awesome-AutoSkill-AutoRubric repo README. |
| argument-hint | Optional: specific topic focus, e.g. "rubric reward modeling" or "skill evolution benchmark" |
| user-invocable | true |
Paper Scout: Daily Auto-Skill & Auto-Rubric Paper Finder
When to Use
- Daily check for new papers in the Auto-Skill / Auto-Rubric domain
- When you want to update the awesome list with recent publications
- When you want a summary of noteworthy new papers on self-evolving agents or rubric learning
Search Topics
Search arxiv for papers matching these keyword groups. Combine multiple queries to maximize coverage.
Auto-Skill Keywords
"self-evolving" agent skill
"skill evolution" LLM agent
"skill creation" agent
"skill library" agent
"skill discovery" agent
"agentic evolution"
"self-improving" agent skill
"harness evolution"
"skill reuse" LLM
Auto-Rubric Keywords
"Auto-Rubric" reward
"rubric" "preference" reward LLM
"rubric learning" alignment
"rubric-based" reward
"rubric generation" LLM
"criteria" "preference" reward modeling
Procedure
Step 1: Search for New Papers
Search arxiv for recent papers using the keyword groups above. Use the fetch_webpage tool to query:
https://arxiv.org/search/?query=KEYWORDS&searchtype=all&order=-announced_date_first
Run multiple searches across both Auto-Skill and Auto-Rubric keyword groups. Focus on papers from the last 7-14 days.
Step 2: Filter and Evaluate
For each candidate paper found:
- Fetch the arxiv abstract page to get full details (title, authors, date, abstract)
- Evaluate relevance — must be directly about:
- Self-evolving agent skills / skill libraries / skill creation / skill evolution, OR
- Rubric learning from preferences / rubric-based reward modeling / auto-rubric generation
- Skip papers that only tangentially mention these topics
Step 3: Present Recommendations
Present findings to the user in this format:
## New Papers Found (Date Range)
### Auto-Skill
1. **Paper Name** (arXiv:XXXX.XXXXX, Date)
- TL;DR: One-sentence summary
- Why it's relevant: Brief note
### Auto-Rubric
1. **Paper Name** (arXiv:XXXX.XXXXX, Date)
- TL;DR: One-sentence summary
- Why it's relevant: Brief note
### Verdict
- 🔥 Must-add: [list]
- 👀 Worth watching: [list]
- ⏭️ Skip: [list with reasons]
Step 4: Update the Repo (upon user confirmation)
After the user confirms which papers to add:
- Read the current README.md at
c:\Users\t-ihchen\Downloads\IHChen\Github\Awesome-AutoSkill-AutoRubric\README.md
- Determine the correct section for each paper:
- Auto-Skill > Skill Learning & Evolution: Papers about skill creation, evolution, reuse
- Auto-Skill > Benchmarks & Evaluation: Benchmarks for skill evaluation
- Auto-Skill > Surveys & Frameworks: Survey papers, frameworks
- Auto-Rubric > Rubric Learning from Preference Data: Papers on learning rubrics from preferences
- Auto-Rubric > Rubric-Guided RL & Reward Modeling: Papers using rubrics for RL/reward
- Auto-Rubric > Rubric-Based Evaluation & Judges: Papers on rubric-based evaluation
- Insert the new entry in chronological order within the section
- Follow the table format:
| **Name** | Date | [arXiv:ID](link) | [Code](link) or - | TL;DR |
- Commit and push:
git add .
git commit -m "Add: Paper1, Paper2, ..."
git push
Important Notes
- Always check if a paper is already in the README before adding
- Use the paper's v1 submission date for the "Date" column
- If a paper has a GitHub repo, include it; otherwise use
-
- Note venue acceptance if mentioned (e.g., ICML 2026, ACL 2026)
- TL;DR should be 1-2 sentences, focusing on the key contribution