菜单
Use for review-only requests for papers or books when the user did not explicitly ask for annotation. Handles the normal single-agent review flow.
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。
基于 SOC 职业分类
| name | paper-review |
| description | Use for review-only requests for papers or books when the user did not explicitly ask for annotation. Handles the normal single-agent review flow. |
| metadata | {"short-description":"Single-agent paper review workflow"} |
Use this skill for the normal single-agent review flow.
Use this skill when the user asks for a review-only pass such as:
Do not use this skill when the user explicitly asks for both annotation and review.
In that case, use annotated-review instead.
If the user explicitly asks for multiple agents, a panel, or a multi-agent review,
use agent-group-discuss instead of this skill.
If the user did not already provide a source path, attached file, PDF, or source tree:
zoterocalibregetscipapers-requesterFor review tasks, do not go online before checking both local libraries.
When you have the document as a local PDF, office file, HTML export, or image-backed scan, prefer docling for structure-aware parsing before relying on ad hoc plain-text extraction.
Use Docling especially when the review depends on:
Before writing review prose, load writing-style-settings.md and record the
active style profile for the review artifact. If the paper is a mathematical,
TCS, graph-theoretic, Lean, or LaTeX manuscript, also load
math-manuscript-style.md. Final review artifacts should record
style_profile_ref, active_overlays, active_requirement_ids, and
style_applied; do not treat a bare style_applied: true assertion as
evidence that the style policy was loaded.
Zotero note storage is off by default for review-only requests.
Do not touch Zotero beyond lookup/retrieval unless the user explicitly asks.
references/common_issues.md and references/reporting_standards.md as internal checklists.source-research/references/specialist-subagents.md without turning the task into a multi-agent run unless the user asked for one.For each issue:
annotated-reviewagent-group-discussWhen this skill is involved, consider these workflow templates (install via
the workflow-templates artifact profile, or --with-deps to pull backing skills):
cross-agent-adversarial-review -- Producer-never-confirmer adversarial review of a paper, proof, or code artifact across agent families with a fresh-agent confirmation gate.Use when the user asks for a multi-agent discussion, panel review, multi-agent review, or multi-agent research session with role selection, round control, and template-based orchestration.
Use when drafting, validating, or normalizing bounded cross-agent task/result packets for parent-controlled handoffs. This is a packet-contract skill, not a runtime delegation broker.
Send email over SMTP using only the Python standard library, with plain-text and HTML bodies, file attachments, cc/bcc, reply-to, a dry-run preview, connection verification, and redacted config inspection.
Runtime helper for autonomous-research-loop ledgers. Use to initialize, append, validate, inspect, or smoke-test autonomous research loop state files without network, package installation, provider CLI calls, or live agent spawning.
Run bounded autonomous research iterations with evidence gates, recovery ledgers, and optional cross-agent handoffs. Use when the user asks to continue research autonomously, run a research loop, integrate autonomous agent loops, or keep improving a research workflow without repeated prompts.
Use for external DOI/ISBN/title resolution, manifest creation from pasted text, and paper retrieval after the local library-first workflow does not satisfy the request or the user explicitly opts out of library use.