help
- Repo stars 330
- License MIT
- Author repo claude-skill-registry
Help (System Assessment)
Deep assessment of workspace state. Identifies blockers, suggests priorities.
When to Activate
- User says: "help", "what should I work on?"
- When stuck or unsure
- Returning after time away
- Multiple things competing for attention
Process
1. Deep Context Check
Git status:
git status -sb
git log --oneline -5
git diff --stat
Inbox:
ls -lt inbox/session-summaries/*.md 2>/dev/null | head -5
ls inbox/ 2>/dev/null
Decisions pending:
ls decisions/*.md 2>/dev/null | head -10
Schema/summary:
cat .claude/tachikoma-summary.yaml 2>/dev/null
2. Report Findings
System Assessment (YYYY-MM-DD):
Git Status:
- Branch: [name]
- Uncommitted changes: [yes/no]
- Recent commits: [summary]
Inbox Status:
- Session summaries: [N files]
- Status: [Clean / Needs processing]
Pending Decisions:
- [N] decisions waiting for review
- [List key ones]
Tachikoma Summary:
- Last scan: [date]
- Observations: [key points]
3. Suggest Priorities
If uncommitted changes:
Priority 1: Git hygiene
- Uncommitted changes in [files]
- Review and commit before other work
If inbox has items:
Priority 1: Process inbox
- [N] session summaries from [dates]
- Run "learn" to update system
- Then "cleanup" to archive
If pending decisions:
Priority 1: Review decisions
- [N] tachikoma decisions pending
- Review and apply or reject
If everything clean:
System is healthy:
- Inbox clean
- Git clean
- No pending decisions
What do you want to work on?
4. Ask What User Wants
Based on this, what do you want to focus on?
Difference from Start/Startup
- start/startup: Quick context, begin work
- help: Deep analysis, identify problems, suggest priorities
Use help when stuck or returning after time away.
- Fluxly category
- Design
- Author-declared agents
- No explicit declaration found; this is not inferred or tested compatibility
- Static check
- 94 / 100 · heuristic scan, not runtime safety proof
- Author / version / license
- @majiayu000 · MIT
- Fluxly token estimate
- Lean
- Fluxly setup estimate
- Plug-and-play
- External API key
- No requirement detected
- Detected OS requirements
- Unspecified
- Runtime requirements
- Unspecified
- Detected file/system behavior
-
- Read-only
- Write / modify
- Detected network behavior
- Local-only
- Install commands
- None (reference only)
Profile is derived at build time from SKILL.md and install vectors. Subject to drift from author intent.
Heads up: 未限定 allowed-tools,默认拥有全部工具权限。
The current SKILL.md does not define a fixed output example. User says: "help", "what should I work on?" When stuck or unsure Returning after time away
Process
Git status: Inbox: Decisions pending:
2. Report Findings
If uncommitted changes: If inbox has items: If pending decisions:
4. Ask What User Wants
# Help (System Assessment)
Deep assessment of workspace state. Identifies blockers, suggests priorities.
## When to Activate
- User says: "help", "what should I work on?"
- When stuck or unsure
- Returning after time away
- Multiple things competing for attention
## Process
### 1. Deep Context Check
**Git status:**
```bash
git status -sb
git log --oneline -5
git diff --stat
```
**Inbox:**
```bash
ls -lt inbox/session-summaries/*.md 2>/dev/null | head -5
ls inbox/ 2>/dev/null
```
**Decisions pending:**
```bash
ls decisions/*.md 2>/dev/null | head -10
```
**Schema/summary:**
```bash
cat .claude/tachikoma-summary.yaml 2>/dev/null
```
### 2. Report Findings
```
System Assessment (YYYY-MM-DD):
Git Status:
- Branch: [name]
- Uncommitted changes: [yes/no]
- Recent commits: [summary]
Inbox Status:
- Session summaries: [N files]
- Status: [Clean / Needs processing]
Pending Decisions:
- [N] decisions waiting for review
- [List key ones]
Tachikoma Summary:
- Last scan: [date]
- Observations: [key points]
```
### 3. Suggest Priorities
**If uncommitted changes:**
```
Priority 1: Git hygiene
- Uncommitted changes in [files]
- Review and commit before other work
```
**If inbox has items:**
```
Priority 1: Process inbox
- [N] session summaries from [dates]
- Run "learn" to update system
- Then "cleanup" to archive
```
**If pending decisions:**
```
Priority 1: Review decisions
- [N] tachikoma decisions pending
- Review and apply or reject
```
**If everything clean:**
```
System is healthy:
- Inbox clean
- Git clean
- No pending decisions
What do you want to work on?
```
### 4. Ask What User Wants
```
Based on this, what do you want to focus on?
```
## Difference from Start/Startup
- **start/startup:** Quick context, begin work
… Author text anchors workflow facts; Fluxly only indexes current sections, terms, files, and commands.
sections -> When to Activate → Process → 1. Deep Context Check → 2. Report Findings → 3. Suggest Priorities → 4. Ask What User Wants
terms -> Git status · Inbox · Decisions pending · Schema/summary · If uncommitted changes · If inbox has items · If pending decisions · If everything clean
files/cmd -> help · inbox/session-summaries · dev/null · Schema/summary · .claude/tachikoma-summary.yaml · yes/no · Start/Startup · start/startup
body sha256 -> 69eac50fd7cf
Decide Fit First
Design Intent
How To Use It
Boundaries And Review