Initial commit

This commit is contained in:
Zhongwei Li
2025-11-30 08:39:39 +08:00
commit 88c4c55b57
5 changed files with 342 additions and 0 deletions

View File

@@ -0,0 +1,15 @@
{
"name": "linear-lifecycle",
"description": "Manage Linear issues via Linearis CLI with zero-context overhead. Create, update, and track issues without loading large MCPs.",
"version": "1.0.0",
"author": {
"name": "Matthew Fornaciari",
"url": "https://github.com/mattforni"
},
"skills": [
"./skills/linear-lifecycle/SKILL.md"
],
"commands": [
"./commands/linear-setup.md"
]
}

3
README.md Normal file
View File

@@ -0,0 +1,3 @@
# linear-lifecycle
Manage Linear issues via Linearis CLI with zero-context overhead. Create, update, and track issues without loading large MCPs.

10
commands/linear-setup.md Normal file
View File

@@ -0,0 +1,10 @@
---
name: linear-setup
description: Configure Linear API token and team key for the linear-lifecycle plugin
---
Copying setup command to clipboard...
```bash
echo "bash ~/.claude/plugins/marketplaces/skillset/plugins/linear-lifecycle/skills/linear-lifecycle/scripts/setup.sh" | pbcopy && echo "✓ Command copied to clipboard! Paste it in your terminal to run the setup."
```

49
plugin.lock.json Normal file
View File

@@ -0,0 +1,49 @@
{
"$schema": "internal://schemas/plugin.lock.v1.json",
"pluginId": "gh:mattforni/skillset:plugins/linear-lifecycle",
"normalized": {
"repo": null,
"ref": "refs/tags/v20251128.0",
"commit": "0e2fd118290016fc48e36030b27c3edc8be30049",
"treeHash": "ffbadf819f26df96d1fa0ab84f6bd9b5dc215b309ee83a18e287e8eadc0ca3af",
"generatedAt": "2025-11-28T10:27:03.114045Z",
"toolVersion": "publish_plugins.py@0.2.0"
},
"origin": {
"remote": "git@github.com:zhongweili/42plugin-data.git",
"branch": "master",
"commit": "aa1497ed0949fd50e99e70d6324a29c5b34f9390",
"repoRoot": "/Users/zhongweili/projects/openmind/42plugin-data"
},
"manifest": {
"name": "linear-lifecycle",
"description": "Manage Linear issues via Linearis CLI with zero-context overhead. Create, update, and track issues without loading large MCPs.",
"version": "1.0.0"
},
"content": {
"files": [
{
"path": "README.md",
"sha256": "be4fe1901701199ba0044ce6baa90c5ff0b0fd7df8f7b4303e1c6bd1aa5351d4"
},
{
"path": ".claude-plugin/plugin.json",
"sha256": "162b6f9f47d9cd4ccf8dcc17e19c10593f4338a7eefe93507bedc6d4b1c9471f"
},
{
"path": "commands/linear-setup.md",
"sha256": "a0e0569b14a9915787cd11e636c6a1a331ab8da8b5a011f40e6fad7355e58212"
},
{
"path": "skills/linear-lifecycle/SKILL.md",
"sha256": "077bd9aa9d1831fd0e84f5bfcc8754d34448d09059ef8a4e9963b427772a8403"
}
],
"dirSha256": "ffbadf819f26df96d1fa0ab84f6bd9b5dc215b309ee83a18e287e8eadc0ca3af"
},
"security": {
"scannedAt": null,
"scannerVersion": null,
"flags": []
}
}

View File

@@ -0,0 +1,265 @@
---
name: linear-lifecycle
description: Use when working with Linear issues across development workflow - uses Linearis CLI with JSON output for zero-context issue management. Get details, create issues, update status, and add comments without consuming tokens in main session.
---
# Linear Lifecycle Management with Linearis CLI
## Overview
**Core principle:** Use Linearis CLI for all Linear operations instead of loading 20k token Linear MCP. CLI returns structured JSON for parsing without context overhead.
**Tool:** [Linearis](https://github.com/czottmann/linearis) - Linear CLI built for LLM agents with ~1000 token footprint vs 13k+ for MCP.
**Context savings:** 100% - no MCP loaded, just bash commands with JSON output.
## Setup
**One-time dependencies install:**
Use `/linear-setup` to install Linearis CLI (required for this skill to work).
**Token configuration:**
Automatic! On first use, Claude will:
1. Check if `~/.linear_api_token` exists
2. If not, prompt you for your Linear API token
3. Save it to `~/.linear_api_token` with secure permissions
4. Verify the connection works
**Get your token:** Linear Settings → Security & Access → Personal API keys
## When to Use
**Use this skill when:**
- Starting work on a Linear issue (need issue details)
- Creating new issues from bugs or features discovered
- Updating issue status during development
- Adding comments or progress updates
- Searching for issues across teams/projects
**Don't use when:**
- Issue tracking not needed for current work
- Working on non-Linear projects
## Implementation
**IMPORTANT: Always check for token on first Linear operation in a skill invocation:**
```bash
# Check if token exists
if [ ! -f ~/.linear_api_token ]; then
echo "⚠️ Linear API token not found."
echo ""
echo "Get your token from: Linear Settings → Security & Access → Personal API keys"
echo ""
# Use AskUserQuestion tool to prompt for token
# Save response to ~/.linear_api_token
# Set permissions: chmod 600 ~/.linear_api_token
# Verify it works with: linearis issues list -l 1
fi
```
After token is confirmed, proceed with Linear operations. Linearis automatically reads from `~/.linear_api_token`.
### Creating a New Issue
**IMPORTANT: Keep it simple! Specify the team key directly. Never use --labels or --priority.**
**User request:** "Create a Linear issue for fixing the avatar crop bug"
**Command:**
```bash
linearis issues create "Fix avatar crop bug" \
--team BET \
--description "Avatar images are cropping incorrectly on mobile devices. Need to adjust aspect ratio handling."
```
**Key rules:**
- ✅ Use `--team TEAM_KEY` (get from user's Linear workspace, e.g. BET, ENG, etc.)
- ✅ Keep description clear and concise
- ❌ NEVER use --labels (causes errors)
- ❌ NEVER use --priority (unnecessary)
**Parse response:**
```bash
# Returns JSON with: {identifier, title, url, ...}
# Extract: issue ID (e.g., BET-145) and URL
```
**Response to user:**
```markdown
✓ Created issue BET-145: Fix avatar crop bug
https://linear.app/your-workspace/issue/BET-145
```
### Starting Work on an Issue
**User request:** "Start working on bet-123"
**Command:**
```bash
linearis issues read BET-123
```
**Parse JSON response for:**
- title
- description
- state (current status)
- priority
- labels
- branchName (suggested git branch)
**Response to user:**
```markdown
Issue: BET-123 - [Title]
Status: [State]
Description: [Brief description]
Labels: [labels]
Branch: [branchName or generate from title]
Creating branch [branch-name]...
```
**Then create branch and proceed with development.**
### Updating Issue Status
**User request:** "Update bet-456 to in progress"
**Command:**
```bash
linearis issues update BET-456 --state "In Progress"
```
**Response:**
```markdown
✓ Updated BET-456 to In Progress
```
### Adding Comments
**User request:** "Add comment to bet-789 about the refactor being done"
**Command:**
```bash
linearis comments create BET-789 --body "Completed auth refactor. Moved from Context API to Zustand for better performance. All tests passing."
```
**Response:**
```markdown
✓ Added comment to BET-789
```
### Searching for Issues
**User request:** "Find all open bugs with label 'authentication'"
**Command:**
```bash
linearis issues search "authentication" --team "$LINEAR_TEAM_KEY" | jq '.[] | select(.labels[]? | contains("bug")) | {id: .identifier, title: .title, state: .state.name}'
```
**Parse and format results as table.**
### Completing Work
**User request:** "Close bet-789, PR merged"
**Commands:**
```bash
# 1. Add completion comment
linearis comments create BET-789 --body "Feature complete. PR #456 merged to main."
# 2. Update status to done
linearis issues update BET-789 --state "Done"
```
**Response:**
```markdown
✓ Marked BET-789 as Done
✓ Added completion comment
```
## Quick Reference
| Operation | Command Pattern |
|-----------|----------------|
| List recent issues | `linearis issues list -l 10` |
| Get issue details | `linearis issues read ABC-123` |
| Create issue | `linearis issues create "Title" --team TEAM_KEY --description "Description"` |
| Update status | `linearis issues update ABC-123 --state "In Progress"` |
| Add comment | `linearis comments create ABC-123 --body "Comment text"` |
| Search issues | `linearis issues search "query"` |
## Common Mistakes
**Not parsing JSON output**
- ❌ Don't show raw JSON to user
- ✅ Parse and format relevant fields cleanly
**Hardcoding team/project names**
- ❌ Don't assume team structure
- ✅ Let user specify or discover via linearis commands
**Using issue IDs incorrectly**
- ❌ Don't lowercase (bet-123) in commands
- ✅ Use proper case (BET-123) - linearis handles both but be consistent
## Real-World Impact
**Before (Linear MCP):**
- 20k tokens consumed at session start
- All tools loaded in context
- Context budget: 180k/200k remaining
**After (Linearis CLI):**
- 0 tokens in session (just bash commands)
- JSON parsing lightweight
- Context budget: 200k/200k remaining
- **100% context savings**
**Performance:**
- Linearis usage docs: ~1000 tokens
- MCP tool definitions: ~13000 tokens
- **92% reduction even for reference material**
## Advanced: Multi-Team Operations
**List issues across teams:**
```bash
linearis issues list --team Frontend -l 5
linearis issues list --team Backend -l 5
```
**Create issue in specific team:**
```bash
linearis issues create "Fix API timeout" --team Backend
```
**No workspace switching needed** - all commands accept `--team` flag for cross-team operations within same Linear workspace.