pipeline-persistence

Pipeline persistence — structured search plan management

Description

Save complex search workflows as reusable Pipeline YAML configurations. Supports:

Quick creation from templates (PICO, comprehensive, exploration, gene_drug)
Custom DAG (directed acyclic graph) multi-step pipelines
Two-tier storage (workspace + global)
Automatic validation + aggressive auto-correction (21 rules)

Trigger Conditions

"Save this search", "Create a search plan"
"Run this search once a week", "Save this pipeline"
"List my pipelines", "Can the last search be run again"
"Convert the search just now into a search plan"

🌟 Quick Start

Method 1: Use a template (simplest)

# 保存一個 PICO 模板 pipeline
save_pipeline(
    name="icu_remimazolam_vs_propofol",
    config="""
template: pico
params:
  P: ICU patients requiring sedation
  I: remimazolam
  C: propofol
  O: delirium incidence, sedation quality
""",
    tags="anesthesia,sedation,ICU",
    description="Weekly monitoring: remimazolam vs propofol ICU sedation"
)

# 執行已保存的 pipeline
unified_search(pipeline="saved:icu_remimazolam_vs_propofol")

Method 2: Custom DAG (full control)

save_pipeline(
    name="brca1_comprehensive",
    config="""
steps:
  - id: expand
    action: expand
    params:
      topic: BRCA1 breast cancer
  - id: pubmed
    action: search
    params:
      query: BRCA1 breast cancer
      sources: pubmed
      limit: 50
  - id: expanded
    action: search
    inputs: [expand]
    params:
      strategy: mesh
      sources: pubmed,openalex
      limit: 50
  - id: merged
    action: merge
    inputs: [pubmed, expanded]
    params:
      method: rrf
  - id: enriched
    action: metrics
    inputs: [merged]
output:
  format: markdown
  limit: 30
  ranking: quality
""",
    tags="genetics,oncology",
    description="BRCA1 breast cancer comprehensive search with MeSH expansion"
)

6 MCP tools

save_pipeline — save a pipeline

save_pipeline(
    name="weekly_remimazolam",     # 唯一名稱 (英數 + _ -, max 64)
    config="<YAML or JSON>",       # Pipeline 配置
    tags="tag1,tag2",              # 逗號分隔標籤
    description="...",             # 人類可讀描述
    scope="auto"                   # "workspace" | "global" | "auto"
)

Auto-correction examples:

action: "find" → auto-corrected to action: "search"(alias resolution)
action: "serach" → auto-corrected to action: "search"(fuzzy matching)
Missing step ID → auto-generate step_1, step_2...
Duplicate step ID → auto-rename s1 → s1_2
Circular dependency → auto-remove problematic references

list_pipelines — list pipelines

list_pipelines()                    # 列出所有
list_pipelines(tag="ICU")           # 按標籤過濾
list_pipelines(scope="workspace")   # 只看工作區

load_pipeline — load a pipeline

load_pipeline(source="weekly_remimazolam")            # 從已保存
load_pipeline(source="file:path/to/pipeline.yaml")    # 從檔案

delete_pipeline — delete a pipeline

delete_pipeline(name="old_search")  # 刪除配置 + 歷史

get_pipeline_history — view execution history

get_pipeline_history(name="weekly_remimazolam", limit=5)
# 顯示：日期、文章數、新增/移除文章、狀態

schedule_pipeline — schedule (Phase 4 not yet implemented)

schedule_pipeline(name="weekly_remimazolam", cron="0 9 * * 1")
# ⚠️ 目前返回使用說明，建議手動執行或使用 OS 排程

4 built-in templates

pico — PICO clinical question

template: pico
params:
  P: ICU patients requiring sedation
  I: remimazolam
  C: propofol
  O: delirium incidence
  sources: pubmed        # 可選，預設 pubmed
  limit: 20              # 可選

Auto-generated DAG:

pico → search_p  ──┐
     → search_i  ──┤
     → search_c  ──┼→ merged → enriched

comprehensive — multiple databases + MeSH expansion

template: comprehensive
params:
  query: CRISPR gene therapy safety
  sources: pubmed,openalex,europe_pmc  # 可選
  limit: 30                             # 可選
  min_year: 2020                        # 可選

Auto-generated DAG:

expand → search_expanded  ──┐
         search_original  ──┼→ merged → enriched

exploration — seed paper exploration

template: exploration
params:
  pmid: "33475315"
  limit: 20        # 每個方向的限制

Auto-generated DAG:

related  ──┐
citing   ──┼→ merged → enriched
refs     ──┘

gene_drug — gene/drug search

template: gene_drug
params:
  term: BRCA1
  sources: pubmed,openalex  # 可選
  limit: 20                  # 可選
  min_year: 2020             # 可選

10 available Actions

Action	Description	Main parameters
`search`	Literature search	`query`, `sources`, `limit`, `min_year`, `max_year`
`pico`	Agent-provided PICO handoff search	`P`, `I`, optional `C`, recommended `O`
`expand`	MeSH/synonym expansion	`topic`
`details`	Get article details	`pmids`
`related`	Related articles	`pmid`, `limit`
`citing`	Citing articles	`pmid`, `limit`
`references`	References	`pmid`, `limit`
`metrics`	iCite citation metrics	(obtained from inputs)
`merge`	Merge results	`method`: `union` / `intersection` / `rrf`
`filter`	Filter results	`min_year`, `max_year`, `article_types`, `min_citations`, `has_abstract`

Two-tier storage model

Workspace scope (.pubmed-search/pipelines/)
├── 每個專案獨立
├── 可納入 git 追蹤
└── 團隊共享

Global scope (~/.pubmed-search-mcp/pipelines/)
├── 跨專案共用
├── 個人偏好
└── 通用模板

Resolution order: workspace first → global fallback

Production-grade examples

Example 1: Weekly report search — anesthesia drug monitoring

name: weekly_anesthesia_monitoring
steps:
  - id: search_remimazolam
    action: search
    params:
      query: remimazolam
      sources: pubmed,europe_pmc
      limit: 50
      min_year: 2024
  - id: search_dex
    action: search
    params:
      query: dexmedetomidine ICU sedation
      sources: pubmed
      limit: 50
      min_year: 2024
  - id: merged
    action: merge
    inputs: [search_remimazolam, search_dex]
    params:
      method: union
  - id: filtered
    action: filter
    inputs: [merged]
    params:
      has_abstract: true
      article_types: "Journal Article,Clinical Trial,Randomized Controlled Trial"
  - id: enriched
    action: metrics
    inputs: [filtered]
output:
  format: markdown
  limit: 30
  ranking: recency

Example 2: Seed paper deep exploration

name: explore_landmark_paper
steps:
  - id: seed_details
    action: details
    params:
      pmids: "33475315"
  - id: related
    action: related
    params:
      pmid: "33475315"
      limit: 30
  - id: citing
    action: citing
    params:
      pmid: "33475315"
      limit: 30
  - id: refs
    action: references
    params:
      pmid: "33475315"
      limit: 30
  - id: merged
    action: merge
    inputs: [related, citing, refs]
    params:
      method: rrf
  - id: enriched
    action: metrics
    inputs: [merged]
output:
  format: markdown
  limit: 25
  ranking: impact

Example 3: Systematic literature review — SGLT2 + heart failure

name: sglt2_heart_failure_review
steps:
  - id: pico
    action: pico
    params:
      P: Type 2 diabetes with heart failure
      I: SGLT2 inhibitors
      C: standard care
      O: hospitalization, mortality
  - id: mesh_expand
    action: expand
    params:
      topic: SGLT2 inhibitors heart failure outcomes
  - id: search_pico_p
    action: search
    inputs: [pico]
    params:
      element: P
      sources: pubmed,europe_pmc
      limit: 100
  - id: search_pico_i
    action: search
    inputs: [pico]
    params:
      element: I
      sources: pubmed,europe_pmc
      limit: 100
  - id: search_expanded
    action: search
    inputs: [mesh_expand]
    params:
      strategy: mesh
      sources: pubmed,openalex
      limit: 100
  - id: merged
    action: merge
    inputs: [search_pico_p, search_pico_i, search_expanded]
    params:
      method: rrf
  - id: filtered
    action: filter
    inputs: [merged]
    params:
      min_year: 2018
      has_abstract: true
  - id: enriched
    action: metrics
    inputs: [filtered]
output:
  format: markdown
  limit: 50
  ranking: quality

Best practices for Agents generating Pipelines

Converting conversation searches into Pipelines

After completing a successful search, the Agent can:

# 1. 回顧剛才的搜尋策略
# 2. 轉化為 Pipeline YAML
save_pipeline(
    name="auto_from_session",
    config="""
steps:
  - id: main_search
    action: search
    params:
      query: "<剛才的查詢>"
      sources: pubmed,openalex
      limit: 50
  - id: enriched
    action: metrics
    inputs: [main_search]
output:
  limit: 20
  ranking: balanced
""",
    description="Auto-generated from search session"
)

# 3. 下次直接重複
# unified_search(pipeline="saved:auto_from_session")

Common patterns

Scenario	Recommended template	Description
Clinical question A vs B	`pico`	Agent provides P/I/C/O; the backend executes a precision/recall search containing O
Topic comprehensive search	`comprehensive`	Multi-source + MeSH expansion
Known important papers	`exploration`	Three-direction exploration (related/citing/refs)
Gene/drug research	`gene_drug`	Vocabulary expansion + multi-source
Complex workflow	Custom DAG	Full control over each step

About this skill

Description

Trigger Conditions

🌟 Quick Start

Method 1: Use a template (simplest)

Method 2: Custom DAG (full control)

6 MCP tools

save_pipeline — save a pipeline

list_pipelines — list pipelines

load_pipeline — load a pipeline

delete_pipeline — delete a pipeline

get_pipeline_history — view execution history

schedule_pipeline — schedule (Phase 4 not yet implemented)

4 built-in templates

pico — PICO clinical question

comprehensive — multiple databases + MeSH expansion

exploration — seed paper exploration

gene_drug — gene/drug search

10 available Actions

Two-tier storage model

Production-grade examples

Example 1: Weekly report search — anesthesia drug monitoring

Example 2: Seed paper deep exploration

Example 3: Systematic literature review — SGLT2 + heart failure

Best practices for Agents generating Pipelines

Converting conversation searches into Pipelines

Common patterns