Install
$ agentstack add skill-makigjuro-cloudstack-ai-plugins-create-tasks ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Create GitHub Tasks
Analyze a feature plan or PRD and create the appropriate GitHub issue structure. Automatically decides whether to create a single issue or an epic with multiple task issues based on scope and complexity.
Configuration
Read cloudstack.json from the project root at the start of execution. Extract relevant fields:
REPO_OWNER/REPO_NAME=repository.owner/repository.name(default: detect fromgit remote -v)
If cloudstack.json does not exist, auto-detect by parsing the GitHub remote URL:
git remote get-url origin | sed -E 's|.*github\.com[:/]([^/]+)/([^/.]+).*|\1 \2|'
Arguments
{prd-path}-- Path to PRD file (optional, will searchdocs/prd/or use conversation context)--dry-run-- Show what would be created without actually creating issues--epic-only-- Create only the epic issue, not individual tasks--no-epic-- Create only task issues, no parent epic
Decision Logic: Single vs Multiple Issues
Create a SINGLE issue when:
- Feature has 3 implementation tasks
- Tasks span multiple services or layers
- A PRD exists with structured sections
- Tasks could be assigned to different people or done in separate PRs
--epic-onlyor--no-epicflags override this logic
Process
Step 1: Find the Plan
- If path provided, read that file
- Else search
docs/prd/*.mdfor most recent PRD - Else look for feature plan in conversation context
- If nothing found, ask user to describe the feature or run
/plan-featureor/prdfirst
Step 2: Parse & Classify
Extract from the plan:
- Title: Feature name
- Summary: Description
- Acceptance Criteria: All AC items
- Tasks: Implementation tasks grouped by layer
- Out of Scope: Exclusions
- Labels: Derive from affected services (see Label Mapping below)
Apply the decision logic above to determine: single issue or epic + tasks.
Step 3: Present Plan for Approval
Show the user what will be created:
## Issue Creation Plan
Mode: {Single Issue | Epic + N Task Issues}
Reason: {brief explanation of why this mode was chosen}
### {Single Issue title OR Epic title}
Labels: {labels}
### Tasks ({count})
1. {task description}
2. {task description}
...
Create? [Yes / Modify / Cancel]
Step 4a: Single Issue Path
Create one issue with all tasks as a checklist using MCP:
mcp__github-mcp-server__create_issue(
owner: "{REPO_OWNER}",
repo: "{REPO_NAME}",
title: "{Feature Title}",
labels: ["{labels}"],
body: "..."
)
Issue body format:
## Summary
{description}
## Acceptance Criteria
- [ ] AC1: ...
- [ ] AC2: ...
## Implementation Tasks
### Domain
- [ ] {task}
### Application
- [ ] {task}
### Infrastructure
- [ ] {task}
### Endpoints
- [ ] {task}
### Frontend
- [ ] {task}
### Tests
- [ ] {task}
## Out of Scope
- {exclusions}
---
Generated with [Claude Code](https://claude.com/claude-code)
Only include layer sections that have tasks. Skip empty sections.
Step 4b: Epic + Tasks Path
Create the epic using MCP:
mcp__github-mcp-server__create_issue(
owner: "{REPO_OWNER}",
repo: "{REPO_NAME}",
title: "Epic: {Feature Title}",
labels: ["epic", "feature", "{labels}"],
body: "..."
)
Epic body format:
## Summary
{summary}
## Acceptance Criteria
{all AC items as checkboxes}
## Task Issues
## Out of Scope
{exclusions}
## References
- PRD: {prd-path if applicable}
---
Generated with [Claude Code](https://claude.com/claude-code)
Create each task issue using MCP:
mcp__github-mcp-server__create_issue(
owner: "{REPO_OWNER}",
repo: "{REPO_NAME}",
title: "[{Layer}] {Task Title}",
labels: ["{layer-label}", "{service-labels}"],
body: "..."
)
Task body format:
## Parent Epic
#{epic-number}
## Task
{task description}
## Acceptance Criteria
{relevant AC items}
## Files to Modify
- `{file path}`
## Implementation Notes
{hints from PRD}
---
Generated with [Claude Code](https://claude.com/claude-code)
Update epic with task links:
mcp__github-mcp-server__update_issue(
owner: "{REPO_OWNER}",
repo: "{REPO_NAME}",
issue_number: {epic-number},
body: "{updated body with - [ ] #{taskN} -- {title} links}"
)
Label Mapping
| Layer/Service | Labels | |---------------|--------| | Domain | domain, backend | | Application | application, backend | | Infrastructure | infrastructure, backend | | Endpoints | api, backend | | Frontend | frontend | | Tests | testing | | Helm/Terraform | infrastructure, devops |
General labels:
featurefor new functionalityenhancementfor improvementsbugfor defects
Add service-specific labels based on the services defined in cloudstack.json or discovered in the project.
Task Title Conventions
Keep titles under 70 characters. Use [Layer] prefix for multi-issue mode:
| Description | Title | |-------------|-------| | "Add user entity to domain" | [Domain] Add user entity | | "Create API endpoint for orders" | [Endpoints] Add order creation endpoint |
Grouping Strategy
Group related tasks into single issues when:
- They modify the same file
- They're logically atomic (must be done together)
- They're trivial (< 10 lines each)
Split into separate issues when:
- Tasks can be done independently
- Different reviewers might handle them
- They touch different services
Output Format
Single Issue Output
## Issue Created
- #{number}: {title}
URL: {url}
## Next Steps
Run `/start-work #{number}` to begin implementation.
Epic + Tasks Output
## Issues Created
### Epic
- #{number}: {title} -- {url}
### Tasks ({count})
| # | Issue | Title | Labels |
|---|-------|-------|--------|
| 1 | #{n1} | {title1} | {labels} |
| 2 | #{n2} | {title2} | {labels} |
## Next Steps
1. Run `/start-work #{first-task}` to begin implementation
2. Close tasks as completed; epic tracks overall progress
Error Handling
- gh not authenticated: Prompt user to run
gh auth login - Label doesn't exist: Create label or skip with warning
- Rate limit: Pause and retry with backoff
- Partial failure: Report which issues were created, which failed
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: makigjuro
- Source: makigjuro/cloudstack-ai-plugins
- License: MIT
- Homepage: https://github.com/makigjuro/cloudstack-ai-plugins#quick-start
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.