Skills for Real Engineers. Straight from my .agents directory.
science-org-roles
Agent skill fromyzlin286-wq/Deepscient
Use when deciding who should do a piece of research work, when a task has no owner or two owners, when a role's authority is unclear (who may approve a design, who may veto an analysis), or when a disagreement must be escalated; covers the 13-role RACI table across S0–S11, the six authority types, the veto bypass to the PI, and the escalation path from shard worker to PI.
How to install science-org-roles
npx skills add yzlin286-wq/Deepscient# or add only this skill to Claude Code: copy the folder
dsh-science/skills/science/science-org-roles/
# into
~/.claude/skills/science-org-roles/ (all projects)
.claude/skills/science-org-roles/ (one project)
1# 科研机构岗位与权限23## 何时用45- 一件事不知道该派谁:检索、设计、跑数、作图、写稿、返修各有人。6- 两方结论冲突,需要知道谁能拍板、谁能否决。7- 一个任务被"大家都可以做"拖住,需要指定唯一负责人。8- 要在 `subagent` 里点名 `agent_preset`,需要确认该岗位对应哪个 preset。910## 四层结构1112| 层 | 名称 | 一句话 | 容量 |13| --- | --- | --- | --- |14| L0 | 课题主持人 | 定方向、拍终稿、对外负责 | 1 |15| L1 | 课题组长 | 拆 Campaign、编排岗位、收 gate | 3 |16| L2 | 岗位专家 | 一个专业面独立作业 | 8–64 |17| L3 | 任务专员 | 执行单个分片,不扩范围 | 20000 |1819容量是该岗位同时可用的逻辑智能体槽位上限,不是建议并发数。实际并发受20`llmConcurrency`(默认 6)与预算共同约束。2122## RACI:13 岗位 × 8 个阶段2324A = 拍板负责,R = 执行,C = 被咨询,I = 被通知,V = 可否决。2526| 岗位 | 选题 | 文献 | 设计 | 执行 | 分析 | 写作 | 内审 | 返修 |27| --- | --- | --- | --- | --- | --- | --- | --- | --- |28| pi | A | I | A | I | I | I | A | A |29| lead | R | A | A | A | A | A | C | A |30| project-manager | C | C | C | C | C | C | C | C |31| survey | C | R | C | I | I | I | I | I |32| strategist | R | C | R | C | I | I | I | I |33| experimenter | I | I | C | R | C | I | I | I |34| analyst | I | I | C | C | R | R | I | I |35| data-steward | I | I | C | C | C | R | C | I |36| methodologist | I | C | V | C | V | C | V | C |37| writer | I | I | I | I | C | R | R | C |38| critic | I | I | C | I | C | C | R | C |39| rebuttal | I | I | I | I | I | I | C | R |40| shard-worker | I | I | I | R | R | R | I | I |4142读法:每个阶段只有一个 A(出现问题找这个人),可以有多个 R。43`project-manager` 在任何阶段都不是 A——它管的是排期,不是内容。4445## 六种权限的边界4647| 权限 | 能做什么 | 不能做什么 |48| --- | --- | --- |49| decide | 在两个可选方案里定一个,决定往下走 | 不能替别的岗位签字,不能改数据 |50| approve | 让一个交付进入下一阶段 | 不能绕过 gate,不能自己给自己 approve |51| review | 出评审意见、分级、要求返工 | 意见必须带位置与改法,否则不算 review |52…
Excerpt shown under the repository's MIT license.
More skills in Deepscient
- ▸science-agent-scheduling
- ▸science-campaign-design
- ▸science-consensus-report
- ▸science-discipline-methods
- ▸science-experiment-runbook
- ▸science-figure-for-journal
- ▸science-frontier-to-paper
- ▸science-fulltext-hunting
- ▸science-journal-intel
- ▸science-journal-selection
- ▸science-literature-search
- ▸science-manuscript-formatting
Similar skill repositories
An agentic skills framework & software development methodology that works.
Public repository for Agent Skills
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
