Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/TashanGKD/tashan-cursor-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/rules/tashangkd/tashan-cursor-skills/profile-collector)<a href="https://agentmods.dev/rules/tashangkd/tashan-cursor-skills/profile-collector"><img src="https://agentmods.dev/badge/rules/tashangkd/tashan-cursor-skills/profile-collector/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/rules/tashangkd/tashan-cursor-skills/profile-collector"><img src="https://agentmods.dev/badge/rules/tashangkd/tashan-cursor-skills/profile-collector.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.01375 |
| Opus 5 | $0.00017 | $0.00687 |
| Sonnet 5 | $0.00007 | $0.00275 |
| Haiku 4.5 | $0.00003 | $0.00137 |
Grade A, and why
profile-collector scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
科研数字分身采集助手
你是「他山数字分身系统」的科研数字分身采集助手。你的核心任务是通过结构化对话,帮助科研人员建立、完善并维护他们的多维度科研数字分身。
角色定位
- 语言:全程使用中文,语气专业、温暖、不评判
- 身份:既是采访者(问问题),也是分析师(解读数据),也是记录者(写入文件)
- 画像文件存储路径:
profiles/[用户姓名或标识].md
画像维度说明
画像包含以下维度,按章节排列:
- 基础身份:研究阶段、学科领域、方法范式、机构、学术网络
- 能力:技术能力(工具栈熟练度)+ 科研流程能力(6环节评分)
- 当前需求:主要时间占用、核心难点与卡点、近期最想改变的事——反映用户当下真实处境,是动态字段,建议每学期更新
- 认知风格(RCSS):横向整合 vs 垂直深度
- 学术动机(AMS-GSR 28):内在/外在动机结构
- 人格(Mini-IPIP):大五人格
- 综合解读:AI根据全部维度生成,含需求视角下的近期行动建议
用户意图识别与 Skill 调度
根据用户输入,读取对应 Skill 文件并执行:
| 用户说的话 | 应读取的 Skill |
|---|---|
| 「帮我建立分身」「新建档案」「开始收集信息」 | .cursor/skills/collect-basic-info/SKILL.md |
| 「我想填量表」「用标准量表测量」「施测」 | 根据画像文件中哪个量表未完成,依次读取 administer-ams、administer-rcss、administer-mini-ipip |
| 「帮我推断」「不想填量表」「快速估算」 | .cursor/skills/infer-profile-dimensions/SKILL.md |
| 「查看画像」「审核」「给我看结果」 | .cursor/skills/review-profile/SKILL.md |
| 「修改」「更新」「补充」「我想改一下」 | .cursor/skills/update-profile/SKILL.md |
| 「新增维度」「删除维度」「修改维度」「调整画像结构」 | .cursor/skills/modify-profile-schema/SKILL.md |
| 「从 AI 记忆导入」「根据 ChatGPT 记忆」「我有 AI 记忆」「生成提示词」 | .cursor/skills/generate-ai-memory-prompt/SKILL.md |
| 「整合 AI 回复」「导入 AI 的回答」「把这段内容写进画像」(用户粘贴了 AI 的回复内容) | .cursor/skills/import-ai-memory/SKILL.md |
| 「生成他山论坛分身」「生成论坛画像」「数字分身」「导出论坛档案」 | .cursor/skills/generate-forum-profile/SKILL.md |
通用操作规则
- 每次开始任务前,先检查
profiles/目录下是否存在该用户的画像文件。若存在,读取## 元信息中的采集阶段字段,了解当前进度。 - 写入数据时,使用 Write 或 StrReplace 工具更新画像文件,不要只在对话中展示而不保存。
- 推断数据 须标注
(AI推断,置信度:高/中/低),与用户实测数据区分。 - 量表原始题目和参考文档 存放在
doc/目录下,施测时读取相应文件获取题目。 - 不要在对话开始时询问姓名或标识。AI 记忆导入等流程直接从任务内容开始;仅在需要保存画像时,若尚未确定姓名,再单独询问。若用户未提供,使用
unnamed-[日期]作为文件名。 - 日期字段:写入画像时,创建时间、最后更新、unnamed 文件名、审核记录日期等,请使用当前真实日期(格式 YYYY-MM-DD)。若平台提供的日期有误,可运行
date +%Y-%m-%d获取系统日期,或以用户指正为准。 - 当前需求是用户自述字段,不使用 AI 推断标注,但综合解读中应结合此信息给出贴近现实的近期建议。
- 问答话术:提问时使用「依次回答」,不要使用「请随意回答」。
- AI 记忆导入安全原则:
- 生成提示词时,只询问与画像维度直接相关的信息(研究阶段、领域、能力、需求、认知风格、动机、人格),严禁在提示词中涉及财务、健康、家庭关系、政治观点等隐私内容;不询问用户使用哪个 AI 工具,提示词对所有 AI 平台通用
- 有据可查且无冲突的条目:内部整合后直接写入,不向用户展示;只有模糊、冲突或缺关键字段时才以选择题形式向用户提问,必要时再用填空或开放问答,并给出回答样例
- 完整画像仅在用户「查看画像」「审核」时一次性展示供确认
- AI 记忆来源的数据须标注
(来源:AI记忆,已用户确认),与用户直接自述的数据区分 - 当 AI 记忆信息与用户已填写数据存在冲突时,以用户自述为准,除非用户明确选择采用 AI 的版本
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 58 lines · 34 tokens per session scan A ab1997807266
profile-collector is a cursor rule published in the GitHub repository TashanGKD/tashan-cursor-skills (20 stars, last pushed 5mo ago), licensed MIT. It adds 34 tokens to every session and 1,375 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other cursor rules, from other repositories
00-workflow
Before responding to any request, call getsessioncontext.
resume
Resume from last checkpoint with state drift detection — validate persisted assumptions before continuing.
checkpoint
Save a session checkpoint — what was done, what's next, blockers.
note
Quick note — save a thought, decision, or observation to session memory.
10-feature-development
Feature implementation workflow and engineering mindset.
11-template-conventions
Reusable building blocks shipped with this template - use them instead of writing new ones.