nexus-lookup
starsforks
How to install nexus-lookup?
niki914/agentic-nexus/nexus-lookup
$npx -y skills add niki914/agentic-nexus --skill nexus-lookupUse when working in the Nexus repository and the task needs project context before code search, editing, debugging, design, or architecture discussion.
SKILL.md
| 1 | # Nexus Lookup |
| 2 | |
| 3 | ## Overview |
| 4 | |
| 5 | This skill is the lookup entry for the Nexus repository. |
| 6 | |
| 7 | Use it to route to the right wiki document before reading code, changing code, debugging behavior, or discussing architecture. The goal is to reduce blind source search and keep project context stable. |
| 8 | |
| 9 | ## When to Use |
| 10 | |
| 11 | Activate this skill when the task involves any of the following: |
| 12 | |
| 13 | - understanding project structure or module boundaries |
| 14 | - locating architecture facts before editing code |
| 15 | - debugging behavior across Xposed entry, hook pipeline, IPC, config, or LLM runtime |
| 16 | - comparing Breeno and XiaoAi behavior |
| 17 | - checking whether a capability is already implemented or still only exists in PRD or task docs |
| 18 | - discussing UI shell, onboarding, settings tree, tool calling, or MCP support |
| 19 | |
| 20 | Do not use this skill for trivial questions that can be answered from one already-known file. |
| 21 | |
| 22 | ## Retrieval Protocol |
| 23 | |
| 24 | Wiki root: `wiki/` |
| 25 | |
| 26 | 1. Read `wiki/index.md` first. |
| 27 | 2. Match the current task against the `检索建议` table. |
| 28 | 3. Read only the recommended documents in order. |
| 29 | 4. Treat child document paths listed in `wiki/index.md` as relative to `wiki/` unless the page explicitly marks them as repo root docs. |
| 30 | 5. Prefer wiki facts first, then inspect source files through the relative paths recorded in the docs. |
| 31 | 6. If wiki and source diverge, trust source code for current implementation and treat design docs as intent only. |
| 32 | 7. Keep the final lookup report concise and continue the actual task after routing is complete. |
| 33 | |
| 34 | ## Output Format |
| 35 | |
| 36 | Use a short lookup report in this shape: |
| 37 | |
| 38 | ```text |
| 39 | 检索报告 |
| 40 | |
| 41 | 已读文档: |
| 42 | - path/to/doc.md — 一句话摘要 |
| 43 | |
| 44 | 相关源码入口: |
| 45 | - relative/path/to/file.kt — 作用 |
| 46 | |
| 47 | 现状判断: |
| 48 | - 已落地 / 进行中 / 提案 |
| 49 | |
| 50 | 下一步建议: |
| 51 | - 建议先读哪个源码或继续做什么 |
| 52 | ``` |
| 53 | |
| 54 | ## Hard Rules |
| 55 | |
| 56 | - Always read `wiki/index.md` before reading any subdocument. |
| 57 | - DO NOT read `docs/.asc_task/` or treat task docs as proof of implementation. |
| 58 | - ALWAYS trust source code for current implementation over design documents. |
| 59 | - Do not paste large code blocks into lookup reports. |
| 60 | - Prefer relative source paths over copied code. |
| 61 | - If an architecture explanation truly requires code excerpts, keep them minimal and explanatory. |
| 62 | |
| 63 | ## Document Tiers |
| 64 | |
| 65 | Use these tiers while reading: |
| 66 | |
| 67 | - `Stable`: stable facts that should match current source structure |
| 68 | - `In Progress`: active design or partially landed work |
| 69 | - `Proposal`: PRD or future-facing intent, not proof of implementation |
| 70 | |
| 71 | ## Common Mistakes |
| 72 | |
| 73 | - jumping directly into grep without reading `wiki/index.md` |
| 74 | - attempting to read or parse `docs/.asc_task/` files instead of relying on actual source code |
| 75 | - treating design documents as the final source of truth instead of trusting the source code |
| 76 | - copying code into wiki instead of pointing to source paths |
| 77 | - reading every document instead of routing by scenario |
People who liked this also viewed
6 picksType
- paperclip-create-agentCreate new agents in Paperclip with governance-aware hiring. Use when you need to inspect adapter configuration options, compare existing agent configs, draft a new agent prompt/config, and submit a hSkillsJul 2026
- subagent-driven-developmentUse when executing implementation plans with independent tasks in the current sessionSkillsJul 2026
- dispatching-parallel-agentsUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesSkillsJul 2026
- browser-act-skill-forgeForges reusable Skill packages (SKILL.md + scripts) from website exploration via browser-act — no re-exploration later. Use when: user wants a reusable Skill for any website, needs to understand a sitSkillsJul 2026
- wizardGenerate an interactive bash wizard that walks a human through a manual procedure — third-party setup, a one-off migration, an A→B state transition — opening URLs, capturing values, confirming each stSkillsJul 2026
- wonda-cliUsing the Wonda CLI to generate images, videos, music, and audio from the terminal — plus LinkedIn, Reddit, and X/Twitter research and automationSkillsJul 2026
