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 agentmods add skills/voidtechnology/voidtech-claude-plugins/researchnpx skills add VoidTechnology/voidtech-claude-plugins --skill researchgit clone --depth 1 https://github.com/VoidTechnology/voidtech-claude-pluginsWhat 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 | $0.00081 | $0.01458 |
| Opus 5 | $0.00041 | $0.00729 |
| Sonnet 5 | $0.00016 | $0.00292 |
| Haiku 4.5 | $0.00008 | $0.00146 |
Grade A, and why
research 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 yesterday.
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 — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research
面向“不够了解,需要先查清楚再决策”的场景。目标不是堆资料,而是把不同来源的证据、分歧、可信度和可执行建议收拢到一份可判断的结论里。
硬约束
- 先界定问题,再搜索;不要把第一个搜索词当成最终研究问题。
- 默认只读:不注册账号、不提交表单、不触发付费动作、不抓取受登录墙或授权限制保护的内容。
- 涉及法律、医疗、金融、安全、合规、价格、政策、版本、公司/人物现状等高变化或高风险事实时,必须使用最新可访问来源,并在结论中写明日期。
- 开放网络内容必须交叉验证。没有两个独立来源支撑的事实,只能标为“单一来源”或“待确认”。
- 不把社区观点当成事实;社区内容用于发现问题、经验和争议,结论仍要回到官方文档、源码、论文、公告、标准或一手数据。
- 输出必须能让用户做下一步决定:给结论、依据、残余风险、推荐动作和需要继续验证的点。
启动检查
- 复述研究问题,并补齐:
- 用户要做的决策是什么。
- 时间范围和地域范围。
- 需要比较的候选方案。
- 成功标准、成本约束和不可接受风险。
- 检查可用工具:
- 若官方
exa插件可用,用于高召回搜索、论文、公司/人物/竞品资料。 - 若官方
firecrawl插件可用,用于抓取单页、站点 crawl、把网页转为干净 Markdown。 - 若官方
youdotcom-agent-skills插件可用,用于带引用的 web research、搜索和内容抽取。 - 若
voidtech-mcp-common已启用,使用 Context7 查询库/框架官方文档;使用 Chrome DevTools 验证需要真实页面行为的主张。
- 若官方
- 如果缺少开放网络工具,仍可继续做“离线研究计划”:基于本地代码、用户提供材料和已知事实输出查询清单、来源优先级和验证步骤,但不要伪造最新结论。
调研分工
优先使用子 agent / Task 工具并行执行;可指定模型时,资料收集 agent 使用 haiku(按量最省、够用),主 agent 保持当前模型做综合判断——不要为资料收集指定更强的模型,这一步的瓶颈是覆盖面而不是推理深度。若当前环境不能委派子 agent,则按相同分工顺序执行并标注为“顺序调研”。
按问题选择 3-5 条资料线(source lane),每条资料线都要给出来源、发现和可信度:
| Lane | 适用来源 | 关注点 |
|---|---|---|
| 官方与一手资料 | 官方文档、发布说明、标准、监管公告、源码、论文、API reference | 当前事实、约束、版本、许可证、使用限制 |
| 实现与维护信号 | GitHub/GitLab issue、PR、commit、release、CI、包仓库 | 活跃度、破坏性变更、真实 bug、维护风险 |
| 社区经验 | 论坛、Reddit、Stack Overflow、博客、HN、Discord 摘要 | 常见坑、迁移成本、非官方 workaround |
| 竞品与替代方案 | 替代库/产品、benchmark、案例、定价页 | 取舍、迁移路径、锁定风险 |
| 本项目适配 | 当前代码库、团队约束、现有 skill/MCP、CI、部署环境 | 落地成本、边界、需要补的验证 |
执行步骤
- 写出 3-7 个搜索查询,覆盖中文、英文、官方名称、常见别名和反向关键词(如
problem、limitations、migration、pricing、security)。 - 为每条 lane 分派任务:
- 要求 agent 返回 5-10 条高价值来源,而不是长篇摘抄。
- 每条来源包含标题、URL、发布日期或更新时间、来源类型、关键发现、可信度和原文可核验位置。
- 明确要求标出“不确定/相互矛盾/过时”的信息。
- 汇总时去重并建立证据表:
事实:多个可信来源一致。强信号:一手来源或大量实现证据支撑。弱信号:社区经验、单一博客、未复现 issue。冲突:来源间结论不一致,说明可能原因。
- 对用户的决策问题给出建议:
- 首选方案和理由。
- 不推荐方案和触发条件。
- 最小验证实验。
- 需要用户或团队人工确认的事项。
输出格式
## 结论
一句话回答用户应该怎么做;如果证据不足,直接说证据不足。
## 依据
| 判断 | 证据 | 可信度 |
|---|---|---|
| ... | 来源与关键发现 | 高/中/低 |
## 分歧与不确定性
- ...
## 建议
1. ...
2. ...
## 下一步验证
- ...
## 来源
- [标题](URL) — 来源类型,日期,为什么使用它
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.
- yesterday First seen · 107 lines · 81 tokens per session scan A b6544f75f161
research is a skill published in the GitHub repository VoidTechnology/voidtech-claude-plugins (2 stars, last pushed 28d ago), licensed Apache-2.0. It adds 81 tokens to every session and 1,458 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…