More screenshots
Works with the AI Agents you already use
Seed Issues: Turn Spec Files Into a GitHub Work Queue
One GitHub issue per spec: verbatim acceptance criteria, risk labels, milestones, dependencies, and never a duplicate.
$5
Seed Issues: Turn Spec Files Into a GitHub Work Queue
Example session with this skill installed
/seed-issues
- Read your context and instructions
- Compiled the seed issues
Summary
Created
11 issues from docs/specs/ (01–11), skipped the overview, labelled 3 as high-risk and wired the dependencies. First unblocked issue:
#1.
| Issue | Spec | Milestone | Labels | Blocked by |
|---|---|---|---|---|
| #1 | 01 Project skeleton | Foundations | spec, area:infra | — |
| #2 | 02 Data model | Foundations | spec, area:backend | #1 |
| #3 | 03 Authentication | Foundations | spec, area:backend, risk:high | #2 |
| #4 | 04 Permissions | Foundations | spec, area:backend, risk:high | #3 |
| #8 | 08 Billing | Features | spec, area:backend, risk:high | #7 |
| #9 | 09 Notifications | Features | spec, area:backend | #8, #6 |
A created issue body (#4, abridged)
Implements `docs/specs/04-permissions.md`.
**Read the spec first - it is the contract. This issue is a pointer to it.**
## Acceptance criteria
- [ ] A user who is not a member of a project gets 404, never 403, for every project URL.
- [ ] Role changes take effect on the next request without re-login.
## Watch for
- 404-vs-403 must be identical in body and headers.
Blocked by #3
Skipped
00-overview.md: a map, not a unit of work (SPEC_SKIP_REGEX).
Next steps
/ship #1or/pickupto start.- Re-run
/seed-issueswhenever
Connects securely to your tools. The creator never sees your data.
About this skill
Agents build well from a queue and badly from a pile of documents. Seed Issues turns your docs/specs/ folder into a GitHub issue queue that any session, human or agent, can work through in order.For every spec it creates one issue with:•the acceptance criteria copied verbatim as checkboxes. A paraphrased criterion is one no reviewer enforces any more.•a "Watch for" list of the rules that can break while tests stay green•spec / area:* labels, a milestone, and risk:high on the specs where a silent failure is most expensive (auth, permissions, money). It proposes that list and asks you before applying it.•Blocked by #N dependencies in build order, plus any cross-links the specs state themselvesSafe to re-run. It matches existing issues by spec path, not title. Add 12-exports.md next month and exactly one new issue appears.Optional Build Runbook. Ask for one and it publishes a private claude.ai page: one card per unit of work, the exact prompt to paste (and which model to use), and a todo / working / done status synced from GitHub. A bundled Perl validator checks the page before and after every update. It refuses to drop an item, reuse an id or publish an older state over a newer one.Pairs with Ship Loop and Pickup. Requires Claude Code and the gh CLI. The runbook also needs perl and claude.ai Artifacts.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 26 days ago
- Accounts
GitHub
- Software
- Claude Code
Git
GitHub CLI
- Optional
- Perl 5.14+claude.ai Artifacts
Sign in to the GitHub CLI. Perl and Artifacts are only for the runbook step; Git for Windows ships Perl.
Permissions
Allowed Hosts
File Scopes
Shell: runs gh to list and create labels, milestones and issues, and perl for the optional runbook validator. Read Files: reads your spec files (SPEC_DIR, default docs/specs) and existing issues. It never edits your specs or source code. Write Files: only .claude/ship-loop.conf (on first run) and scratch copies of the runbook page outside the repository. Network/GitHub: creates labels, milestones and issues; it lists all existing issues first and skips any spec that already has one, so it never duplicates. It asks you before applying risk:high labels it inferred. Optional runbook: publishes or updates one private claude.ai page.