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.
npx skills add cuic19053-hue/-skills- --skill social_surveygit clone --depth 1 https://github.com/cuic19053-hue/-skills-Wrote 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/skills/cuic19053-hue/-skills-/social_survey)<a href="https://agentmods.dev/skills/cuic19053-hue/-skills-/social_survey"><img src="https://agentmods.dev/badge/skills/cuic19053-hue/-skills-/social_survey/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/skills/cuic19053-hue/-skills-/social_survey"><img src="https://agentmods.dev/badge/skills/cuic19053-hue/-skills-/social_survey.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 | $0.00084 | $0.15034 |
| Opus 5 | $0.00042 | $0.07517 |
| Sonnet 5 | $0.00017 | $0.03007 |
| Haiku 4.5 | $0.00008 | $0.01503 |
Grade A, and why
social_survey 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 12d 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 — 966 lines — stays where its author put it; the contents beside it link to each section on GitHub.
三下乡社会实践-调研类 Skill
本 skill 处理"大学生暑期三下乡社会实践-调研类"立项申报书全流程：信息采集 → 按天方案撰写 → 问卷设计 → docx 生成 → 质检。仅处理"调研类"子类（走访调研 + 问卷 + 深度访谈为主），支教/政策宣讲/科技服务类请走对应 skill。
输出文件：可提交的 .docx（A4 纸张，宋体小四正文，黑体三号标题，1.5 倍行距，首行缩进 2 字符）。
一、适用场景与触发条件
适用：用户提到"三下乡""社会实践""调研类""走访调研""问卷调查""深度访谈""暑期社会实践立项申报书"等关键词，且实践主要形式为走访村庄 + 问卷发放 + 深度访谈（非支教、非政策宣讲、非科技服务）。
不适用（需走其他 skill）：
- 支教类（含课时安排、教学对象为中小学生）→ 走支教 skill
- 政策宣讲类（含宣讲场次、宣讲对象为村民）→ 走宣讲 skill
- 科技服务类（含技术培训、设备维修）→ 走服务 skill
- 单纯志愿服务（无调研产出）→ 走志愿服务 skill
触发后第一步：确认实践形式为"调研类"（含问卷 + 访谈）。若用户未明确，按调研类处理并告知，避免与支教/宣讲混入。
调研类 vs 其他子类核心区别：
- 调研类：核心产出是"调研报告"（1.5~2 万字），用数据说话，强调方法科学性
- 支教类：核心产出是"教学课时 + 学生反馈"
- 宣讲类：核心产出是"宣讲场次 + 受众覆盖"
- 服务类：核心产出是"服务记录 + 物资捐赠"
二、工作流程总览
四阶段串行执行，每阶段完成后再进入下一阶段：
阶段 1 信息采集（10~15 轮对话）： 按"团队信息 → 实践主题 → 选址理由 → 调研对象 → 抽样方法 → 样本量 → 问卷维度 → 实施按天安排 → 安全保障 → 预期成果 → 经费预算"顺序采集。每轮不超过 5 个字段。调研专属字段必须追问清楚，禁止编造样本量与访谈对象。
阶段 2 内容撰写（一次性产出）： 按 12 栏目顺序撰写。实施方案严格按天列出，每天标注走访村庄 + 访谈人数。问卷设计明确维度与题型。预期成果以调研报告为主。
阶段 3 docx 生成：
调用 python build.py --data data.json --out output.docx 生成 Word 文档。data.json 字段定义见第十二章。
阶段 4 质检： 按第十三章清单逐项检查。任何一项不达标返回阶段 2 修改。
禁止行为：
- 跳过采集直接套模板
- 用"开展调研活动""进行走访"等空泛表述
- 编造样本量、访谈对象姓名、村庄数据
- 把支教/宣讲内容混入调研类申报书
- 用"若干""一定""相关"等模糊词
三、信息采集清单
3.1 通用必采（10 字段）
| 字段 | 说明 | 示例 | 缺失追问策略 |
|---|---|---|---|
| team_name | 团队名称（含赴 + 地点 + 主题） | 赴 XX 县乡村振兴调研团 | "团队正式名称？" |
| theme | 实践主题（对应团中央年度主题） | 乡村振兴 | "对应团中央哪一年度主题？" |
| location | 实践地点（精确到村） | XX 省 XX 县 XX 镇 5 个行政村 | "精确到哪几个村？" |
| practice_time | 实践时间（含出发/返程） | 2025.07.15-07.21（7 天） | "出发与返程日期？" |
| team_size | 团队人数（6~15 人） | 10 人 | "团队几人？建议 6~15 人" |
| leader_name | 队长姓名 | 张三 | "队长全名？" |
| leader_id | 队长学号 | 202212345 | "学号？" |
| leader_phone | 队长联系电话 | 138XXXXXXXX | "手机号？" |
| college | 申报单位（学院全称） | 经济管理学院 | "学院全称？" |
| apply_date | 申报日期 | 2025 年 5 月 20 日 | "申报提交日期？" |
3.2 调研专属必采（核心）
| 字段 | 说明 | 示例 |
|---|---|---|
| research_object | 调研对象 | XX 县 5 个行政村 18~65 岁常住村民 |
| sampling_method | 抽样方法 | 分层随机抽样（按村分层，村层内简单随机） |
| sample_size | 样本量 | 384 份（n=Z²σ²/E²，Z=1.96，σ=0.5，E=0.05） |
| questionnaire_dims | 问卷维度 | 5 维度 25 题（人口学 4+经济 6+教育 5+医疗 5+文化 5） |
| questionnaire_scale | 量表类型 | 李克特 5 级量表（11 题）+ 单选 8 题 + 多选 6 题 |
| interview_targets | 访谈对象 | 8 位村干部 + 3 位致富带头人 + 5 位普通村民 |
| interview_outline | 访谈提纲 | 4 个一级问题，每个含 3~4 个追问点 |
| reliability_check | 信效度检验 | 预调研 30 份，Cronbach α ≥ 0.7 后正式发放 |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 966 lines · 84 tokens per session scan A bedcad6fd997
social_survey is a skill published in the GitHub repository cuic19053-hue/-skills- (10 stars, last pushed 1mo ago), licensed MIT. It adds 84 tokens to every session and 15,034 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
foundry-hosted-agent-validation
Step-by-step process for validating a Python Foundry hosted agent sample (under python/samples/04-hosting/foundry-hosted-agents/) end to end — running it locally (native runtime and azd ai agent run) and after deploying it to an Azure AI Foundry project with azd. Use this when asked to validate a hosted agent sample.
skill-doc-delivery
Convert markdown to DOCX, PPTX, XLSX, PDF office documents — use when you need exportable deliverables.
document-direct-python
Use direct Python execution for reliable document creation including spreadsheets, PDFs, and structured reports.
parse-document
Convert a PDF, scan, image of a page, or office file to clean markdown through the connected Superlinked MCP edge, so the source document is not read into model context directly. Use when the user asks to read, parse, OCR, extract from, summarize, or answer questions about a document.
pdf-text-extraction-fallback-85d5ca
Fallback workflow for extracting text from PDFs when readfile returns binary data.
pdf-verification-cli
Verify PDF page count and content using command-line tools when Python libraries unavailable.