A research assistant that searches arXiv, a public archive of research papers, for recent work on self-improving AI agents, reusable agent skills, and automated rubric or preference learning.

Use it for daily searches, updates to the Awesome-AutoSkill-AutoRubric list, and summaries of noteworthy papers on agent evolution, skill libraries, rubric learning, reward modeling, or preference alignment.

Finding relevant new papers across several overlapping topics is time-consuming. This narrows the search and identifies papers worth considering for an Awesome list or research update.

paper-scout — Skill for Claude Code by AaronCIH

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add AaronCIH/Awesome-AutoSkill-AutoRubric --skill paper-scout

Clone the repo

git clone --depth 1 https://github.com/AaronCIH/Awesome-AutoSkill-AutoRubric

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/aaroncih/awesome-autoskill-autorubric/paper-scout/github.svg)](https://agentmods.dev/skills/aaroncih/awesome-autoskill-autorubric/paper-scout)

Your own site

<a href="https://agentmods.dev/skills/aaroncih/awesome-autoskill-autorubric/paper-scout"><img src="https://agentmods.dev/badge/skills/aaroncih/awesome-autoskill-autorubric/paper-scout/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/aaroncih/awesome-autoskill-autorubric/paper-scout"><img src="https://agentmods.dev/badge/skills/aaroncih/awesome-autoskill-autorubric/paper-scout.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 72 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,056 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00072	$0.01056
Opus 5.5	$0.00029	$0.00422
Sonnet 5.5	$0.00014	$0.00211
Haiku 4.5	$0.00007	$0.00106

Measured 2mo ago against content hash 7a05d1678c6e, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

paper-scout scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/skills/paper-scout/SKILL.md · 111 lines

How it starts

The opening of the file, as written. The whole thing — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.

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]

Read the full file on GitHub · 111 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2mo ago First seen · 111 lines · 72 tokens per session scan A 7a05d1678c6e

Subscribe to this mod's changes

paper-scout is a skill published in the GitHub repository AaronCIH/Awesome-AutoSkill-AutoRubric (9 stars, last pushed 5mo ago), licensed MIT. It adds 72 tokens to every session and 1,056 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Building agents Preference Alignment Research Reward Modeling agent-skills auto-skills harness paper-with-code

paper-scout next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
paper-scoutAaronCIH/Awesome-AutoSkill-AutoRubric	Skill	9	5mo ago	MIT	A	72	Claude Code
open-code-reviewalibaba/open-code-review	Skill	50k	yesterday	Apache-2.0	A	98	Cursor
google-drive-sheetsyc-software/qm	Skill	16k	today	MIT	A	32	Claude Code, Codex
connect-appsyc-software/qm	Skill	16k	today	MIT	A	22	Claude Code, Codex
interactive-loginyc-software/qm	Skill	16k	today	MIT	A	46	Claude Code, Codex
migrating-openai-agents-sdk-to-pydantic-aipydantic/pydantic-ai	Skill	21k	today	MIT	A	87	Codex
last30Dayschaitanyagiri/munder-difflin	Skill	9.3k	6d ago	MIT	A	83	Claude Code