Agent Auto Sci Automation
科研自动化与长期记忆子 skill。用于把 AutoSci/OmegaWiki 的论文摄入、知识库、idea 生成、novelty 检查、实验设计、失败经验、论文计划、review/rebuttal、可视化和进化记录流程转化为本地 Codex 科研工作流。触发于科研自动化、跨项目记忆、source manifest、wiki式知识库、paper ingestion、idea pipeline、experiment lifecycle、checkpoint、API 配置等任务。
Install
$ agentstack add skill-lzy599775-agent-auto-sci-skills-agent-auto-sci-automation ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Agent Auto Sci Automation
Use this subskill to convert a one-off research task into a persistent workflow.
Fast Workflow
- Define the project root, source folders, output folders, and ownership boundaries.
- Create or update a source manifest: papers, notes, data, scripts, figures, manuscripts.
- Convert sources into structured memory: tables, indexes, concept notes, method notes, failed attempts.
- Route the task to literature, idea, experiment, analysis, writing, review, or archive stages.
- Add checkpoints for long-running workflows.
- Update the evolution archive after workflow improvements.
Read references/autosci_workflow_and_memory.md.
For deeper AutoSci encapsulation:
references/autosci_command_mapping.md: local mapping of AutoSci/init,/ideate,/novelty,/exp-design,/paper-plan,/review,/rebuttal, and visualization flows.references/api_and_checkpoint_templates.md: source manifest, checkpoint, API role, and failure-memory templates.
Default Memory Entities
sources: PDFs, data, notes, web pages.papers: structured paper records, DOI, topic, method, findings.concepts: definitions, boundaries, related concepts.methods: reusable analytical or writing methods.ideas: proposed, rejected, validated, or failed research ideas.experiments: empirical or computational tests, including failed runs.outputs: manuscripts, figures, tables, posters, slides.reviews: reviewer comments, internal critiques, response strategies.evolution: skill changes, new workflows, validation evidence.
Quality Rules
- Checkpoint before long workflows.
- Do not overwrite user-curated source boundaries without explicit instruction.
- Record failed ideas and weak evidence, not only successful outputs.
- Do not write secrets into logs or HTML.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Lzy599775
- Source: Lzy599775/agent-auto-sci-skills
- License: MIT
- Homepage: https://lzy599775.github.io/agent-auto-sci-skills/
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.