Agent Skills › ymx10086/ResearchClaw › research-workflows

research-workflows

GitHub

管理以项目为中心的研究工作流、声明、证据和实验。当任务结果需持久化存储至结构化研究图谱而非仅作为临时对话回复时，使用此技能。

src/researchclaw/agents/skills/research_workflows/SKILL.md ymx10086/ResearchClaw

触发场景

用户进行长期研究项目且需保存成果需要将研究数据（如声明、实验）持久化到结构化系统中

安装

npx skills add ymx10086/ResearchClaw --skill research-workflows -g -y

更多选项

非标准路径

npx skills add https://github.com/ymx10086/ResearchClaw/tree/main/src/researchclaw/agents/skills/research_workflows -g -y

不安装直接使用

npx skills use ymx10086/ResearchClaw@research-workflows

指定 Agent (Claude Code)

npx skills add ymx10086/ResearchClaw --skill research-workflows -a claude-code -g -y

安装 repo 全部 skill

npx skills add ymx10086/ResearchClaw --all -g -y

预览 repo 内 skill

npx skills add ymx10086/ResearchClaw --list

SKILL.md

Frontmatter

{
    "name": "research-workflows",
    "emoji": "🧭",
    "triggers": [
        "workflow",
        "research workflow",
        "project dashboard",
        "claim evidence",
        "experiment tracking",
        "structured notes"
    ],
    "description": "Manage project-centric research workflows, claims, evidence, notes, and experiments. Use when the task should be persisted into ResearchClaw's structured research graph instead of remaining a transient chat reply."
}

Research Workflows

Use this skill when the user is working on a long-running project and the result should be persisted into the structured research system.

Tools

research_projects_list
research_project_create
research_project_dashboard
research_workflows_list
research_workflow_create
research_workflow_get
research_workflow_add_task
research_workflow_update_task
research_workflow_tick
research_note_create
research_notes_search
research_claim_create
research_claim_attach_evidence
research_claim_graph
research_experiment_log
research_experiment_compare

Guidance

Prefer storing durable work in the structured graph when it will matter beyond the current turn.
Use research_workflow_get before updating a workflow task.
When making a substantive claim, attach evidence instead of leaving it implicit.
When logging an experiment, include metrics or archived outputs whenever possible.