xiaohongshu
xiaohongshu
22-understand-deep-150
[22] UNDERSTAND. Deep research from all sources — internal (code, git, logs) AND external (web, docs, best practices). Use when choosing libraries, investigating solutions, understanding legal/technical questions, comparing approaches, or anytime you need comprehensive knowledge from both project...
Full skill instructions
Core Principle: Research deeply from ALL sources — internal context AND external knowledge. Don't limit yourself to one world.
When you invoke this skill, you're researching from TWO worlds:
🏠 INTERNAL (Project Context)
🌍 EXTERNAL (World Knowledge)
🏠 INTERNAL (Project Context) 🌍 EXTERNAL (World Knowledge)
├── Code files ├── Web search
├── Git history/blame ├── Official documentation
├── Comments & inline docs ├── Stack Overflow / forums
├── Config files ├── Blog posts & tutorials
├── Tests (expected behavior) ├── GitHub issues/discussions
├── Logs (actual behavior) ├── Best practice guides
└── Project documentation └── Comparison articles
Key Insight: Most questions need BOTH worlds. Internal tells you "what we have", external tells you "what's possible".
| Scenario | Internal Research | External Research |
|---|---|---|
| Choosing a library | What's already used? Conflicts? | Comparisons, benchmarks, community health |
| Legal question | Context, contracts, history | Laws, regulations, precedents |
| Architecture decision | Current structure, constraints | Patterns, case studies, trade-offs |
| Bug investigation | Code, logs, git blame | Known issues, similar problems |
| Performance optimization | Profiling, bottlenecks | Best practices, techniques |
| Security concern | Current implementation | Vulnerabilities, OWASP, CVEs |
| Question | Best Source | Tools* |
|---|---|---|
| "What does this code do?" | Read the code | read_file, grep |
| "Why was it written this way?" | Git blame/history | git log, git blame |
| "When did this change?" | Git history | git log --follow |
| "What was the original intent?" | Commits + PR descriptions | git show, git log |
| Question | Best Source | Tools* |
|---|---|---|
| "How do others solve X?" | Web search | web_search |
| "What's the best practice?" | Web + official docs | web_search |
| "Is there a library for X?" | Web + package repos | web_search |
| "What are common pitfalls?" | Web + Stack Overflow | web_search |
| Question | Best Source | Tools* |
|---|---|---|
| "What does this API do?" | Official documentation | web_search for docs |
| "What version is compatible?" | Package docs + changelogs | web_search, read_file |
| "How has this evolved?" | Git history | git log, git diff |
*Tool names vary by agent — use equivalent available tools
Different agents have different tool names. Use what's available:
| Capability | Claude Code | Codex | Generic |
|---|---|---|---|
| Read files | read_file | view_file | file reading tool |
| Search code | grep, codebase_search | search | code search tool |
| Web search | web_search | search_web | internet search tool |
| Run commands | run_terminal_cmd | shell | terminal/shell tool |
| Git operations | terminal + git | terminal + git | git via terminal |
Rule: Use the tool that achieves the goal, regardless of its name.
Clarify the knowledge gap:
Choose appropriate sources based on need:
Need to understand existing code?
→ Read code + git history + comments
Need to find a solution approach?
→ Web search + best practices + similar projects
Need to debug something?
→ Logs + git blame + error searches
Need to integrate with external system?
→ Official docs + web search + examples
Minimum 3 sources for important facts:
Check for consistency:
Combine findings into verified knowledge:
Git history is often overlooked but invaluable:
| Situation | Git Command | What You Learn |
|---|---|---|
| "Why is this code weird?" | git blame <file> | Who wrote it and when |
| "What was the intent?" | git log -p <file> | Changes over time |
| "When did this break?" | git bisect | Which commit caused issue |
| "What changed recently?" | git log --since="1 week" | Recent modifications |
| "How did this function evolve?" | git log -p -S "function_name" | History of specific code |
# Who wrote this line and why?
git blame path/to/file.js
# What changed in this file and why?
git log -p --follow path/to/file.js
# Find when a specific line was added
git log -S "specific code" --oneline
# What happened around a date?
git log --since="2024-01-01" --until="2024-02-01"
Web search is essential for:
"How to [specific task] in [language/framework]"
"[error message] solution"
"[library name] best practices"
"[problem] vs [alternative] comparison"
"[technology] production tips"
When using Research-Deep 150:
🔬 **Research-Deep 150 Complete**
**Question:** [What we needed to know]
**🏠 Internal Research:**
- 📁 Code: [what we found in codebase]
- 📜 Git: [what history revealed]
- 📋 Context: [relevant project facts]
**🌍 External Research:**
- 🌐 Web: [what we found online]
- 📖 Docs: [official documentation]
- 💡 Best Practices: [established patterns]
**Synthesis:**
[Combined understanding from BOTH worlds]
**Recommendation:** [What to do based on research]
**Confidence:** [X]%
**Sources agree:** [Yes/Partial/No]
**Gaps:** [What we still don't know]
| Level | Sources | Confidence |
|---|---|---|
| Strong | 3+ agreeing sources, primary data | 90-100% |
| Good | 2-3 sources, mostly agree | 75-89% |
| Moderate | 1-2 sources, some assumptions | 50-74% |
| Weak | Single source or contradictions | <50% |
Question: "Which date library should we use?"
AI: "Use moment.js, it's popular" (outdated advice)
Result: Chose deprecated library, had to migrate later
Question: "Which date library should we use?"
🔬 Research-Deep 150:
🏠 Internal Research:
- 📁 Code: Currently using native Date, some formatting issues
- 📋 Context: Need timezone support, bundle size matters
- 📜 Git: Previous attempt with moment.js reverted (too heavy)
🌍 External Research:
- 🌐 Web: "best javascript date library 2024"
- 📖 Docs: Compared date-fns, dayjs, luxon
- 💡 Best Practices: moment.js deprecated, modern alternatives exist
Synthesis:
- moment.js is deprecated (external confirms)
- We tried it before, too heavy (internal confirms)
- date-fns: modular, tree-shakeable, good for bundle size
- dayjs: moment-like API, small size
- luxon: best for complex timezone work
Recommendation: dayjs for simple needs (2KB), date-fns for complex (modular)
Confidence: 92%
Question: "Can we use this image in our product?"
🔬 Research-Deep 150:
🏠 Internal Research:
- 📁 Found image in /assets/hero.jpg
- 📜 Git: Added 2 years ago, no license info in commit
- 📋 Context: Used on marketing page
🌍 External Research:
- 🌐 Reverse image search: Found on Unsplash
- 📖 Unsplash license: Free for commercial use, no attribution required
- 💡 Best Practice: Still good to keep license record
Recommendation: Safe to use. Document source in assets README.
Confidence: 95%
| Failure | Detection | Recovery |
|---|---|---|
| Internal only | No external perspective | Add web search, docs |
| External only | Ignoring project context | Check code, git, configs |
| Guessing | No sources cited | Stop, research properly |
| Outdated info | Old articles, deprecated advice | Check dates, find current sources |
| One world bias | Only looking at one side | Deliberately check the other world |
| Skill | Focus |
|---|---|
| goal-clarity-150 | WHAT we need to know |
| research-deep-150 | WHERE to find answers (internal + external) |
| deep-think-150 | HOW to reason about findings |
| impact-map-150 | WHAT the findings affect |
After completing this skill, write to .sessions/SESSION_[date]-[name].md:
### [HH:MM] Understand-Deep 150 Complete
**Topic:** <research topic>
**Sources:** <internal/external sources used>
**Key Findings:** <bullet points>
**Synthesis:** <combined insight>
Remember: Most questions need BOTH worlds. Internal context tells you "what we have and why". External knowledge tells you "what's possible and what others learned". Deep research combines both for complete understanding.
xiaohongshu
technical spec
product ux expert
database patterns
Conduct multi-agent task orchestration and workflow coordination.
Initialize project with Conductor artifacts (product definition,
Expert in web animations, transitions, and motion design using Framer Motion and CSS
Creates Mermaid and ASCII diagrams for flowcharts, architecture, ERDs, state machines, mindmaps, and more. Use when user mentions diagram, flowchart, mermaid, ASCII diagram, text diagram, terminal diagram, visualize, C4, mindmap, architecture diagram, sequence diagram, ERD, or needs visual docume...
PostgreSQL bindings for H3 hexagonal grid system. Use when working with H3 cells in Postgres, including spatial indexing, geometry/geography integration, and raster analysis.
Context-Driven Development skill for projects using Conductor. Use this skill when you detect a `conductor/` directory in the project, when working on tasks defined in a `plan.md` file, or when the user asks about tracks, specs, or plans. Automatically applies TDD workflow, tracks task completion...
Display project status, active tracks, and next actions
Official Stakpak application containerization standard operating procedure, a step-by-step guidline to properly dockerize applications. This is a rule book curated by the Stakpak Team.
Generate, edit, and beat-sync AI video with leading models in one workspace.
The world's fastest calendar for remote work
Transform Your Design with AI Designer by ImgCreator.ai
Revolutionizing Video Production with AI-Powered Creativity
Extend an image past the frame and let AI fill the new aspect ratio.
Discover your celebrity doppelgänger with StarByFace!
ChainClarity explains 700+ crypto whitepapers in plain English, with layered summaries, comparisons, research tools, alerts, and a $4.99 Pro plan.
Opus.ai: Revolutionize Your Web Experience