Academic Pipeline — Academic 流水线

Academic 流水线 — 10-Stage End-to-End (Hermes Edition)

Orchestrates the full academic 工作流 from re搜索 idea to publication-ready 输出.

10 Stages Stage 技能 Used Key 输出

• Planning intake + mode selection Project Plan with stage gating
• Re搜索 deep-re搜索 (full/quick) RQ Brief + Methodo记录y + Bibliography + Synthesis
• Paper Draft academic-paper (full) Complete paper draft
• Peer Review academic-paper-reviewer (full) 5-panel review + Editorial Decision
• Revision academic-paper (revision) Revised draft + 追踪ed changes
• Polish academic-paper (revision) Final polished manuscript
• Ethics 检查 academic-paper-reviewer (quick) Ethics clearance
• AI Disclosure academic-paper (disclosure) Venue-specific AI 状态ment
• 格式化 academic-paper (格式化-convert) LaTeX/DOCX/PDF + cover letter
• Deliver file 输出 Publication package

The 编排器 (mAIn Hermes 代理):

Stage 1: Clarify user intent → select modes → 创建 Project Plan Stage 2-6: Sequential delegate_task calls, each loading the relevant 技能 Stage 7-9: Lightweight parallel or sequential based on dependencies Stage 10: Write final package to disk Key 检查points Stage 1→2: User 应用roves Project Plan Stage 2→3: Re搜索 报告 accepted by user Stage 4→5: Review decision 接收d Stage 6→7: Polished draft ready 恢复 Support

流水线 状态 stored in ~/.hermes/会话s/. User can 恢复 with:

Continue the academic 流水线 from Stage 4

Usage Patterns

Full 流水线:

Take "AI impact on HE 质量 assurance" through the full 流水线 to publication-ready manuscript

Partial 流水线:

启动 from Stage 3 with my existing deep-re搜索 输出

恢复:

恢复 流水线 from Stage 5 with reviewer comments: [comments]

Integration with Hindsight Memory

流水线 状态 and key 输出s are retAIned to Hindsight for long-term 追踪ing:

hindsight memory retAIn --content "[stage_输出]" --tags "流水线,stage-N"

Critical Rules ⚠️ Each stage MUST complete before the next begins ⚠️ User confirmation at all IRON RULE 检查points ⚠️ Ground truth isolation: review stage must NOT see author 身份