project-onboard
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chineseproject-onboard
project-onboard
Reads the current project's file system and determines which of 6 onboarding cases applies, then recommends the exact command sequence.
Triggers: , onboard project, diagnose project state, what do I run first, project setup help
/project-onboard读取当前项目的文件系统,判断适用6种项目初始化场景中的哪一种,然后推荐精准的命令序列。
触发词:/project-onboard、onboard project、diagnose project state、what do I run first、project setup help
Process
流程
I run a 5-check waterfall in strict priority order. I read real files — I never ask the user questions. The first failing check determines the primary case assignment. Check 4 is non-blocking: a project can simultaneously be Case 6 (healthy) and have local skill issues.
我会按照严格的优先级顺序执行5项检查。我会读取真实文件——绝不会向用户提问。首次未通过的检查将确定主要场景分配。检查4为非阻塞项:项目可同时属于场景6(健康状态)并存在本地skill问题。
Check 1 — CLAUDE.md exists
检查1 — CLAUDE.md是否存在
Read (also accept at project root for global-config repos).
.claude/CLAUDE.mdCLAUDE.mdIf absent → Case 1: Brand-new project
undefined读取.claude/CLAUDE.md(对于全局配置仓库,也接受项目根目录下的CLAUDE.md)。
若不存在 → 场景1:全新项目
undefinedDiagnosis
Diagnosis
Project state: Case 1 — Brand-New Project
Detected:
- .claude/CLAUDE.md: NOT FOUND
- No Claude configuration present in this project
Warnings:
- None
Project state: Case 1 — Brand-New Project
Detected:
- .claude/CLAUDE.md: NOT FOUND
- No Claude configuration present in this project
Warnings:
- None
Recommended Command Sequence
Recommended Command Sequence
- /project-setup — creates .claude/CLAUDE.md, ai-context/ skeleton, persists context to engram
- /memory-init — generates ai-context/ files from real project content
- /project-audit — produces audit-report.md with score and findings
- /project-fix — applies all corrections from the audit report
- /project-setup — creates .claude/CLAUDE.md, ai-context/ skeleton, persists context to engram
- /memory-init — generates ai-context/ files from real project content
- /project-audit — produces audit-report.md with score and findings
- /project-fix — applies all corrections from the audit report
Notes
Notes
After /project-fix, re-run /project-audit to verify score ≥ 75 and SDD Readiness = FULL or PARTIAL.
See ai-context/scenarios.md → Case 1 for failure modes and recovery steps.
Stop here if Case 1.
---After /project-fix, re-run /project-audit to verify score ≥ 75 and SDD Readiness = FULL or PARTIAL.
See ai-context/scenarios.md → Case 1 for failure modes and recovery steps.
若为场景1,流程到此结束。
---Check 2 — Engram MCP reachable
检查2 — Engram MCP是否可达
Check if Engram MCP is reachable (call ).
mem_context- If Engram is reachable → SDD infrastructure lives in Engram. Treat the project as healthy for this check and continue to Check 3.
- If Engram is NOT reachable → Case 2 (below).
If Engram not reachable → Case 2: CLAUDE.md present but no SDD persistence
undefined检查Engram MCP是否可达(调用)。
mem_context- 若Engram可达 → SDD基础设施存储于Engram中。本次检查判定项目健康,继续执行检查3。
- 若Engram不可达 → 场景2(如下)。
若Engram不可达 → 场景2:存在CLAUDE.md但无SDD持久化
undefinedDiagnosis
Diagnosis
Project state: Case 2 — CLAUDE.md Without SDD Persistence
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: NOT REACHABLE
- SDD cannot persist artifacts without Engram
Warnings:
- [list any ai-context/ files found or note if ai-context/ is absent]
Project state: Case 2 — CLAUDE.md Without SDD Persistence
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: NOT REACHABLE
- SDD cannot persist artifacts without Engram
Warnings:
- [list any ai-context/ files found or note if ai-context/ is absent]
Recommended Command Sequence
Recommended Command Sequence
- Ensure Engram MCP server is running and configured
- /project-audit — diagnose the full scope of what is missing
- /project-fix — adds SDD section to CLAUDE.md
- /memory-init — if ai-context/ is empty or absent
- /project-audit — verify score improved
- Ensure Engram MCP server is running and configured
- /project-audit — diagnose the full scope of what is missing
- /project-fix — adds SDD section to CLAUDE.md
- /memory-init — if ai-context/ is empty or absent
- /project-audit — verify score improved
Notes
Notes
project-fix will ask before every change — review each proposed action carefully.
See ai-context/scenarios.md → Case 2 for failure modes and recovery steps.
Stop here if Case 2.
---project-fix will ask before every change — review each proposed action carefully.
See ai-context/scenarios.md → Case 2 for failure modes and recovery steps.
若为场景2,流程到此结束。
---Check 3 — ai-context/ has ≥ 3 populated files
检查3 — ai-context/包含≥3个已填充内容的文件
Read directory. Count files that exist AND have more than 10 lines.
Expected files: , , , , .
ai-context/stack.mdarchitecture.mdconventions.mdknown-issues.mdchangelog-ai.mdIf fewer than 3 populated files → Case 3: Partial SDD, sparse memory layer
undefined读取目录。统计存在且内容超过10行的文件数量。预期文件包括:、、、、。
ai-context/stack.mdarchitecture.mdconventions.mdknown-issues.mdchangelog-ai.md若已填充内容的文件少于3个 → 场景3:SDD不完整,内存层稀疏
undefinedDiagnosis
Diagnosis
Project state: Case 3 — Partial SDD (ai-context/ is sparse)
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/ populated files: [N] of 5 (minimum needed: 3)
- Missing or empty: [list each absent/stub file]
Warnings:
- [list any other issues found, e.g. stale onboarding.md]
Project state: Case 3 — Partial SDD (ai-context/ is sparse)
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/ populated files: [N] of 5 (minimum needed: 3)
- Missing or empty: [list each absent/stub file]
Warnings:
- [list any other issues found, e.g. stale onboarding.md]
Recommended Command Sequence
Recommended Command Sequence
- /memory-init — regenerates all ai-context/ files from real project state
- /project-audit — verify D2 score improved
- /project-fix — address any remaining findings
- /memory-init — regenerates all ai-context/ files from real project state
- /project-audit — verify D2 score improved
- /project-fix — address any remaining findings
Notes
Notes
/memory-init does not overwrite files that already have substantial content.
See ai-context/scenarios.md → Case 3 for failure modes and recovery steps.
Stop here if Case 3.
---/memory-init does not overwrite files that already have substantial content.
See ai-context/scenarios.md → Case 3 for failure modes and recovery steps.
若为场景3,流程到此结束。
---Check 4 — Local skills review (non-blocking)
检查4 — 本地skill检查(非阻塞)
Read directory. If it exists and has any subdirectories, record a warning for the diagnosis output. Do not stop — continue to Check 5.
.claude/skills/Local skills warning (append to any case diagnosis):
Warnings:
- .claude/skills/ found with [N] local skill(s): [list names]
Run /project-audit to see Dimension 9 findings (duplicate detection, structural completeness, language compliance).
Run /project-fix Phase 5 to apply corrections after auditing.读取目录。若该目录存在且包含子目录,在诊断输出中记录警告。不要停止流程——继续执行检查5。
.claude/skills/本地skill警告(追加至任意场景诊断结果):
Warnings:
- .claude/skills/ found with [N] local skill(s): [list names]
Run /project-audit to see Dimension 9 findings (duplicate detection, structural completeness, language compliance).
Run /project-fix Phase 5 to apply corrections after auditing.Check 5 — Orphaned SDD changes
检查5 — 孤立的SDD变更
Search engram for active SDD changes: . Filter for changes that have a state artifact but no archive-report, and check if they have tasks and verify-report artifacts.
mem_search(query: "sdd/", project: "{project}")If any change is missing OR artifacts → Case 5: Orphaned or stale changes
tasksverify-reportundefined在engram中搜索活跃的SDD变更:。筛选存在状态工件但无归档报告的变更,并检查它们是否包含任务和验证报告工件。
mem_search(query: "sdd/", project: "{project}")若任意变更缺少或工件 → 场景5:孤立或过期变更
tasksverify-reportundefinedDiagnosis
Diagnosis
Project state: Case 5 — Orphaned SDD Changes
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/: adequate ([N] populated files)
- Orphaned changes:
- [change-name]: missing [tasks | verify-report]
- [change-name]: missing [tasks | verify-report]
Warnings:
- [local skills warning if Check 4 triggered]
Project state: Case 5 — Orphaned SDD Changes
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/: adequate ([N] populated files)
- Orphaned changes:
- [change-name]: missing [tasks | verify-report]
- [change-name]: missing [tasks | verify-report]
Warnings:
- [local skills warning if Check 4 triggered]
Recommended Command Sequence
Recommended Command Sequence
- /sdd-status — see all active changes and their current phase
- For each orphaned change, one of: /sdd-apply <name> — if tasks.md is missing (change never implemented) /sdd-verify <name> — if implemented but never verified /sdd-archive <name> — if complete but never archived
- /project-audit — verify D3 shows no orphaned changes
- /sdd-status — see all active changes and their current phase
- For each orphaned change, one of: /sdd-apply <name> — if tasks.md is missing (change never implemented) /sdd-verify <name> — if implemented but never verified /sdd-archive <name> — if complete but never archived
- /project-audit — verify D3 shows no orphaned changes
Notes
Notes
To discard a dead-end change: run /sdd-archive <name> to close it in engram.
See ai-context/scenarios.md → Case 5 for failure modes and recovery steps.
Stop here if Case 5.
---To discard a dead-end change: run /sdd-archive <name> to close it in engram.
See ai-context/scenarios.md → Case 5 for failure modes and recovery steps.
若为场景5,流程到此结束。
---Check 6 — All healthy (default)
检查6 — 全部健康(默认场景)
All checks passed. The project is fully configured.
undefined所有检查均通过。项目已完全配置。
undefinedDiagnosis
Diagnosis
Project state: Case 6 — Fully Configured
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/: adequate ([N] populated files)
- Engram: no orphaned SDD changes
Warnings:
- [local skills warning if Check 4 triggered]
- [stale onboarding.md warning if Last verified > 90 days: "Run /project-update to refresh user docs"]
Project state: Case 6 — Fully Configured
Detected:
- .claude/CLAUDE.md: FOUND
- Engram MCP: REACHABLE
- ai-context/: adequate ([N] populated files)
- Engram: no orphaned SDD changes
Warnings:
- [local skills warning if Check 4 triggered]
- [stale onboarding.md warning if Last verified > 90 days: "Run /project-update to refresh user docs"]
Recommended Command Sequence
Recommended Command Sequence
For a well-understood change:
- /sdd-propose <change-name> — create proposal (orchestrator fast-forwards planning automatically)
- /sdd-apply <change-name> — implement the task plan
For a complex or vague change:
- /sdd-explore <topic> — investigate the area before committing to a change
- /sdd-propose <change-name> — create proposal from exploration findings
- /sdd-apply <change-name> — implement the task plan
For a well-understood change:
- /sdd-propose <change-name> — create proposal (orchestrator fast-forwards planning automatically)
- /sdd-apply <change-name> — implement the task plan
For a complex or vague change:
- /sdd-explore <topic> — investigate the area before committing to a change
- /sdd-propose <change-name> — create proposal from exploration findings
- /sdd-apply <change-name> — implement the task plan
Notes
Notes
Multi-phase flows (propose → spec+design → tasks in one shot, or full cycles with exploration) are handled
by the orchestrator as meta-commands typed directly in conversation — not invoked as skills.
See ai-context/quick-reference.md for entry-point guidance.
---Multi-phase flows (propose → spec+design → tasks in one shot, or full cycles with exploration) are handled
by the orchestrator as meta-commands typed directly in conversation — not invoked as skills.
See ai-context/quick-reference.md for entry-point guidance.
---Check 7 — Runtime sync hint (non-blocking, global-config mode only)
检查7 — 运行时同步提示(非阻塞,仅全局配置模式)
If AND exist at the project root (i.e., the cwd is the meta-repo), append the following line to the section of the case diagnosis output — regardless of which case was assigned:
install.shskills/agent-configWarnings:- Run /claude-folder-audit to verify ~/.claude/ is in sync with this repo (installation drift check).This check MUST NOT change the case assignment, alter the Recommended Command Sequence, or interrupt the existing 6-check waterfall. It is a hint only.
若项目根目录同时存在和(即当前工作目录为元仓库),则在场景诊断结果的部分追加以下内容——无论分配的是哪个场景:
install.shskills/agent-configWarnings:- Run /claude-folder-audit to verify ~/.claude/ is in sync with this repo (installation drift check).本检查不得更改场景分配、修改推荐命令序列或中断现有6项检查的流程。它仅作为提示信息。
Stale docs warning logic
过期文档警告逻辑
After determining the primary case, check , , and for the field. If any file's date is more than 90 days from today, append to the Warnings section:
ai-context/onboarding.mdai-context/scenarios.mdai-context/quick-reference.md> Last verified:- [filename] is stale ([N] days since last verification). Run /project-update to refresh.确定主场景后,检查、和中的字段。若任意文件的日期距离当前日期超过90天,在Warnings部分追加:
ai-context/onboarding.mdai-context/scenarios.mdai-context/quick-reference.md> Last verified:- [filename] is stale ([N] days since last verification). Run /project-update to refresh.Rules
规则
- Do not ask the user any questions — detection is fully automatic from file-system state
- Detection is derived from real file presence/absence — no hardcoded case IDs or lookup tables
- Priority order is strict: once a case is assigned at check N, do not also assign a lower-priority case (exception: Check 4 local skill flag is always surfaced as a warning, non-blocking)
- Emit structured output only — no raw directory listings, no stack traces, no internal file dumps
- Make no file-system changes — this skill is 100% read-only
- Always include a section pointing to the relevant case in
## Notesfor failure modesai-context/scenarios.md
- 不得向用户提问——完全通过文件系统状态自动检测
- 检测基于真实文件的存在/缺失——无硬编码场景ID或查找表
- 优先级顺序严格:一旦在检查N分配了场景,不得同时分配低优先级场景(例外:检查4的本地skill标记始终作为警告显示,非阻塞)
- 仅输出结构化内容——无原始目录列表、无堆栈跟踪、无内部文件转储
- 不得修改文件系统——本skill为100%只读
- 始终包含部分,指向
## Notes中相关场景的故障模式说明ai-context/scenarios.md