# artemiz-linear

> Use when working with Linear issues (ART-XX, issue identifiers), tracking work items, creating session notes, managing project context, or when user mentions Linear, projects, sprints, cycles, or issue tracking. Triggers on: "work on ART-", "linear", "session note", "track this", issue identifiers, project names. Commands: /linear-load, /linear-track, /linear-save, /linear-focus, /linear-verify.

- Author: Nate Grabowski
- Repository: NateGrabowski/artemiz-linear
- Version: 20260125032008
- Stars: 1
- Forks: 0
- Last Updated: 2026-02-06
- Source: https://github.com/NateGrabowski/artemiz-linear
- Web: https://mule.run/skillshub/@@NateGrabowski/artemiz-linear~artemiz-linear:20260125032008

---

---
name: artemiz-linear
description: >
  Use when working with Linear issues (ART-XX, issue identifiers), tracking work items,
  creating session notes, managing project context, or when user mentions Linear, projects,
  sprints, cycles, or issue tracking. Triggers on: "work on ART-", "linear", "session note",
  "track this", issue identifiers, project names. Commands: /linear-load, /linear-track,
  /linear-save, /linear-focus, /linear-verify.
---

# Artemiz Linear Skill

Claude + Linear workflow enforcement for predictable, reliable behavior.

## Current Context (auto-loaded)

!`npx tsx scripts/get-context.ts 2>/dev/null || echo "Run /linear-load to initialize"`

## Commands

| Command | Purpose |
|---------|---------|
| `/linear-load [project]` | Switch projects or refresh context |
| `/linear-track [title]` | Create issue from detected work |
| `/linear-save` | Create session note as Project Update |
| `/linear-focus <issue>` | Deep-dive on specific issue |
| `/linear-verify <issue>` | Gap analysis before closing |

## Core Behaviors

### On "Work on ART-XX"

When user says "work on [issue]":
1. If no context loaded, run `/linear-load` first
2. Load issue with `/linear-focus`
3. Show relationships: blocks, blocked-by, related
4. Update status to In Progress
5. Begin work with full context

### Issue Detection

Recognize trackable items during conversation:
- Bug reports ("this is broken", "found a bug")
- Feature ideas ("we should add", "it would be nice if")
- Technical debt ("need to refactor", "this is messy")
- Tasks taking >15 minutes

**Behavior:**
1. Identify the item
2. Check for similar existing issues
3. Consider blocking relationships
4. Suggest: "Want me to track that? Related to ART-XX"
5. On confirmation, create in Triage status

### Session Notes

**Proactively suggest** `/linear-save` at:
- Session end
- Before context switch
- After 2+ hours of work
- At natural breakpoints
- After completing an issue

Say: "Good stopping point - want me to create a session note?"

**Template (strict):**
```markdown
## Date
YYYY-MM-DD

## What I worked on
- [bullet points]

## Where I left off
- [current state]

## Next steps when resuming
1. [numbered steps]

## Key Insights
- [discoveries]

## Questions
- [open items]
```

**Health auto-detection:**
- `onTrack` - Good progress, no blockers
- `atRisk` - Blockers encountered, questions pending
- `offTrack` - Stuck, major issues

### Gap Analysis Before Closing Issues

Before marking any issue Done, **use sequential thinking** for systematic verification:

```
npx tsx scripts/gap-analysis.ts ART-XX
```

**5-Step Process (use sequential thinking to work through each):**
1. **Deliverables** - List all from description (checkboxes, acceptance criteria)
2. **Verification** - Confirm artifacts exist
3. **Related Issues** - Check sub-issues complete, update blocked issues
4. **Scope Creep** - Track discovered work (TODOs, "should also")
5. **Unblocking** - What does completing this enable?

**Automatic:** Claude performs this (with sequential thinking) before any status change to Done.
**Manual:** User can run `/linear-verify ART-XX` anytime.

### Multi-Step Work

Always use TaskCreate/TaskUpdate for work with 2+ steps:
- Create tasks at start
- Update status as you work
- Mark completed when done

### Documentation Location

Default to Linear unless explicitly asked for local file:
- Workflow docs, decisions, ADRs -> Linear
- Code docs (README, API) -> Local file
- Session notes -> Project Update

### Sub-issue vs Edit vs Comment

```
New work discovered?
├── <15 min, part of current work? -> Edit description
├── Needs separate tracking? -> Create sub-issue
└── Just context/information? -> Add comment
```

| Action | When |
|--------|------|
| Sub-issue | New work, own "done" criteria, >15 min |
| Edit | Spec change, acceptance criteria update |
| Comment | Progress, decisions, blockers, discoveries |

## Tool Reliability

MCP tools may timeout. Fallback to scripts when needed.

| Operation | Primary | Fallback |
|-----------|---------|----------|
| Get issue | MCP `get_issue` | - |
| Create issue | MCP `create_issue` | - |
| Update status | MCP `update_issue` | `scripts/lib/linear-api.mjs` |
| Add comment | MCP | `scripts/lib/linear-api.mjs` |
| Project update | - | `scripts/lib/linear-api.mjs` |

## References

For detailed specifications:
- **Full ADR decisions**: See `references/contract.md`
- **Command details**: See `references/commands.md`
- **Content hierarchy**: See `references/content-hierarchy.md`