/wrap
Session end - document updates, commit
$ npx -y skills add mag123c/toktrack --skill wrap --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/wrap
Context preview
The summary Claude sees to decide when to auto-load this skill.
Session end - document updates, commit
SKILL.md
wrap.SKILL.mdname: wrap
description: Session end - document updates, commit
required_context:
- .claude/ai-context/architecture.md
Wrap
Flow
Git Status → Doc Check → User Selection → Execute → Complete
Execution
1. **Git Status**
git status --short
git diff --stat HEAD~3
2. **Doc Check Checklist (MUST)** | Change | Target | Required | |--------|--------|----------| | trait/type | architecture.md | on structural change | | convention | conventions.md | on new pattern | | **Decisions** | **.dev/DECISIONS.md** | **on new feature/design** |
**DECISIONS.md entries (required check)**:
- [ ] New feature implemented → record decision rationale, alternatives, reasons
- [ ] Divergence between PLAN and actual implementation → add `**Implementation Note**`
- [ ] New pattern/convention → also update conventions.md
3. **Plan Reconciliation (MUST)**
Reconcile the PLAN's provenance tags/gates (`[unverified-gate]`/`[agent-inferred]`/Phase 0) one-by-one — **close (confirmed/falsified/changed) or explicitly delegate (delegated). No silent TODOs.** Unsettled gates = wrap incomplete.
Detailed procedure·delegated block·escalation criteria (WRAP=detail/DECISIONS=why/CLAUDE.md=critical·recurring) → `references/reconciliation.md`
4. **User Selection**: AskUserQuestion 5. **Execute**: Run selected items
DSL Rules (ai-context)
- Table > prose
- Codeblock > description
- Core only, minimize lines
Commit
{type}({scope}): {summary}Completion
wrap complete = **skill chain finished** Next task starts with new `/clarify` or `/next`
Read more
name: wrap description: Session end - document updates, commit required_context: - .claude/ai-context/architecture.md
Wrap
Flow
Git Status → Doc Check → User Selection → Execute → Complete
Execution
1. **Git Status**
git status --short git diff --stat HEAD~3
2. **Doc Check Checklist (MUST)** | Change | Target | Required | |--------|--------|----------| | trait/type | architecture.md | on structural change | | convention | conventions.md | on new pattern | | **Decisions** | **.dev/DECISIONS.md** | **on new feature/design** |
**DECISIONS.md entries (required check)**:
- [ ] New feature implemented → record decision rationale, alternatives, reasons
- [ ] Divergence between PLAN and actual implementation → add `**Implementation Note**`
- [ ] New pattern/convention → also update conventions.md
3. **Plan Reconciliation (MUST)**
Reconcile the PLAN's provenance tags/gates (`[unverified-gate]`/`[agent-inferred]`/Phase 0) one-by-one — **close (confirmed/falsified/changed) or explicitly delegate (delegated). No silent TODOs.** Unsettled gates = wrap incomplete.
Detailed procedure·delegated block·escalation criteria (WRAP=detail/DECISIONS=why/CLAUDE.md=critical·recurring) → `references/reconciliation.md`
4. **User Selection**: AskUserQuestion 5. **Execute**: Run selected items
DSL Rules (ai-context)
- Table > prose
- Codeblock > description
- Core only, minimize lines
Commit
{type}({scope}): {summary}Completion
wrap complete = **skill chain finished** Next task starts with new `/clarify` or `/next`
Ultra-fast token & cost tracker for LLM Token Usage (e.g. Claude Code)
Other skills on toktrack.
- /clarify
Adaptive requirements clarification with auto-depth routing. Shallow (Q&A) for simple tasks, Deep (exploration + DRAFT + PLAN) for complex ones. Escalates automatically when ambiguity persists.
Open skill - /implement
TDD implementation (RED→GREEN→REFACTOR) → verify → review
Open skill - /next
Session start - check progress, suggest next task
Open skill - /review
Multi-agent code review for Rust CLI/TUI. UX Review inactive (terminal UI — no web frontend). Code Review only. Includes Rust/clippy-specific checklist.
Open skill - /verify
Self-healing verification loop (test → clippy → fmt)
Open skill

