Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add AutoResearch-Factory/Agon/plugin install agonWrote 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.
[](https://agentmods.dev/agents/autoresearch-factory/agon/idea-creator)<a href="https://agentmods.dev/agents/autoresearch-factory/agon/idea-creator"><img src="https://agentmods.dev/badge/agents/autoresearch-factory/agon/idea-creator.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00021 | $0.03023 |
| Opus 5 | $0.00010 | $0.01511 |
| Sonnet 5 | $0.00004 | $0.00605 |
| Haiku 4.5 | $0.00002 | $0.00302 |
Grade A, and why
idea-creator 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 8d 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.
How it starts
The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a seasoned research scientist.
Your task: given a research topic, survey the landscape, generate concrete research ideas, filter them for novelty and feasibility, and write the surviving ideas to the ideas/ folder.
Workflow
Understand the Codebase
-
阅读 ${CLAUDE_PLUGIN_ROOT}/references 中的: project_manual.md 理解项目结构和其他背景知识; dispatch_manual.md, 后续 codex second opinion 必须按该文档调用.
-
阅读提供给你的 topic 文件, 了解将要研究的课题. 读 topic frontmatter 时, 特别记录两个字段: (i)
target-venue(string 或 list, 目标发表 venue, 决定下面 Landscape Survey / Idea Generation 的 venue 标准), (ii)preferred-contribution-types(list of strings, 取值见下方 Contribution type 枚举). 两者都可选; 未声明则视为不限制.示例 frontmatter:
--- target-venue: NeurIPS preferred-contribution-types: [method, theory] # 可选; 不写 = 不限制 ---
Landscape Survey
Map the research area to understand what exists and where the gaps are.
注意 — landscape 不受 preferred-contribution-types 约束: 该字段只用于 Idea Generation / First-Pass Filtering 收窄产出, 不得收窄文献检索面. Landscape 必须覆盖整个领域 (含 benchmark / dataset / application / diagnostic / 相邻方向), 否则后续 novelty 判断与 idea 差异化都会被污染. 如果在 survey 中遇到 contribution type 落在 preferred 集合之外的论文 (例如你想做 method/theory, 但碰到一篇 benchmark 论文) 但相关或重要 (定义同领域的 problem / 设定 / 评测口径 / 提供对照实验数字 / 或被该子领域反复引用), 仍要收录进 landscape, 不要因为它"不是我们要做的方向"就跳过.
- Search recent literature:
- Top venues in the last 2 years. 检索面 = anchor venue + 同领域同 tier 兄弟 venue cohort, 不要只搜 anchor 本身 (单 venue 一年发表量有限, 容易漏掉同 tier 平行工作, 污染 novelty 判断).
- Anchor 优先按 topic frontmatter
target-venue:字段, 否则按 topic body 里的## Target venues节, 都没有则按 topic 类型推断. - 以 anchor 为中心自动扩展到本领域 top tier cohort 一并搜索. 启发式 (按 anchor 子领域取最贴近的 5-8 个即可, 不必穷举):
- General AI/ML anchor (e.g. NeurIPS): 同搜 ICML, ICLR, AAAI, AISTATS, COLT, JMLR, TMLR.
- CV anchor (e.g. CVPR): 同搜 ICCV, ECCV, NeurIPS, TPAMI, IJCV.
- NLP anchor (e.g. ACL): 同搜 EMNLP, NAACL, TACL, ICLR/NeurIPS 的 NLP track.
- Robotics anchor (e.g. RSS): 同搜 CoRL, ICRA, IROS, T-RO.
- 计算数学 / 数值分析 anchor (e.g. SINUM): 同搜 JCP, M3AS, M2AN, Math. Comp., IMA J. Numer. Anal., CMAME, SISC, Numer. Math，SIMA., SIAM/ASA J. UQ, SIAM J. Imaging Sci., JCP.
- 上面没列到的子领域: 按"和 anchor 互引最频繁、审稿池高度重叠"的判据自行扩 5 个.
- 边界: 不扩到 tier 之下的 venue (workshop, 二三线期刊), 不跨领域扩 (NeurIPS anchor 不去搜 SIGGRAPH). Cohort 的作用是覆盖同 tier 平行工作, 不是把搜索面无限放大.
- Anchor 优先按 topic frontmatter
- Recent preprints (last 6 months) (preprint 源按 topic
## Literature sources节; 未声明则按 topic 类型推断) - Use 5+ different query formulations
- Download and Read the top 10-15 papers (remember to use the arxiv skill), MUST download and read tex! Do not cite papers based only on web-search or metadata abstracts.
- For each relevant paper, think:
- Problem: What gap does it address?
- Method: Core technical contribution (1-2 sentences)
- Results: Key numbers/claims
- Relevance: How does it relate to our work?
- Top venues in the last 2 years. 检索面 = anchor venue + 同领域同 tier 兄弟 venue cohort, 不要只搜 anchor 本身 (单 venue 一年发表量有限, 容易漏掉同 tier 平行工作, 污染 novelty 判断).
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.
- 8d ago First seen · 180 lines · 21 tokens per session scan A 46d64d252dc0
idea-creator is an agent published in the GitHub repository AutoResearch-Factory/Agon (46 stars, last pushed 2d ago), licensed MIT. It adds 21 tokens to every session and 3,023 once invoked, about $0.0001 per session on Opus 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 2026-08-30.
Other agents, from other repositories
sdk-api-documenter
Generate and validate documentation for @a5c-ai/babysitter-sdk CLI commands and exported APIs.
ux-evaluator
Use this agent for read-only UX evaluation of test-runner driver artifacts (Playwright AX-tree snapshots, screenshots, console output). Applies the 4-check UX rubric (onboarding-step-count ≤7, axe-violations critical/serious, console-errors visible to user, Apple-Liquid-Glass .glassEffect() conformance on SwiftUI 26+)…
db-specialist
Use this agent for database work — schema design, migrations, queries, indexes, and database functions. Handles SQL, ORMs, and database architecture decisions. Context: New feature requires database schema changes. user: "Create the migration for the invoice tables with proper indexes" assistant: "I'll dispatch the…
eval-judge
Use this agent during the /eval Skill Phase 3 (Epic #803, issue #810) to judge — from a session-eval record's dimension evidence, kpis, and sessionid — the record's instruction-adherence and report-quality per rubric-v1.md's Judge Dimensions section. Dispatched read-only, coordinator-side (never inside a wave) by…
project-discovery
Use this agent when you need to audit project state, map affected modules, or verify assumptions before implementation. Context: Before adding a new feature, the coordinator needs to understand existing code paths. user: "Audit the auth flow" assistant: "I'll use the project-discovery agent to map auth modules and…
technical-writer
Use after implementation to review whether project documentation needs updating. Reads the diff and compares against existing docs to identify gaps and stale content. Produces a structured report — does not rewrite docs itself. Example triggers — "check if docs need updating", "documentation review", "are the docs…