菜单
Produce a research / analysis / deep-dive report as a disposable artifact. No SDLC pipeline, no evaluator, no security review.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/cwijayasundara/claude_harness_eng_v5 --skill research命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Code generation quality principles — TDD, typing, error handling, logging, API integration, LLM integration.
Use when creating or editing .claude/agents/*.md, .claude/skills/*/SKILL.md, or .claude/commands/* — apply docs/prompting-standards.md (trigger conditions, model-agnostic bodies, no reasoning-as-text, effort floors).
Autonomous build loop with Karpathy ratcheting, GAN evaluator, and session chaining. Iterates story groups until all features pass or stopping criteria met.
| name | research |
| description | Produce a research / analysis / deep-dive report as a disposable artifact. No SDLC pipeline, no evaluator, no security review. |
Produces a cited research or analysis write-up. Disposable artifact lane — no pipeline, no agents teams, no ratchet gates, no security review.
Prefer deep-research. If the deep-research skill is available, invoke it — it fans out web searches, fetches sources, adversarially verifies claims, and synthesizes a cited report. Before invoking, make sure the question is specific enough; if it is underspecified, ask 2–3 clarifying questions to narrow scope, then pass the refined question.
If deep-research is not installed, run a focused research pass yourself: search, read sources, cross-check key claims against more than one source, and synthesize. Cite sources inline.
Write the report to research/ (or a path the user names) as Markdown and report the path. This is an artifact for human consumption — it does not need tests or a verification gate.