- Home
- Skills
- Code Quality & Review
- Team Agent Skills: Ticket to Merge-Ready
Team Agent Skills: Ticket to Merge-Ready
Readiness is a comment ("🟢 PASS — Ready for human review"), never a GitHub approval, so an automated verdict can never enable the merge button.
Free
Works with the AI Agents you already use
About this skill
A team of agent roles (Developer, Reviewer and Manual QA) that takes a ticket from idea to merge-ready. The roles work in separate sessions and coordinate only through signed pull-request comments. Readiness is a comment ("🟢 PASS — Ready for human review"), never a GitHub approval, so an automated verdict can never enable the merge button. No skill edits a ticket: everything is posted as an append-only comment.
Included skills
| Skill | What it does |
|---|---|
| tickets | Shows what's on your plate from live data: open PRs with the next move on each, merged work not yet closed, and what you can pick up or is blocked. |
| create-clarity | Checks an issue against the real code and posts an evidence-backed specification before anyone builds it. |
| implement | Takes one issue to one open pull request in an isolated git worktree, with local verification and review. |
| pr-review | Reviews one PR once against its ticket, then publishes validated findings and a verdict. |
| pr-fix | Works one PR's outstanding feedback and failing checks once as the Developer. |
| manual-qa | Tests a PR like a user in an isolated local environment and publishes evidence. |
| watch-and-review | Re-runs pr-review on every new revision of a PR. |
| watch-and-fix | Re-runs pr-fix whenever a PR owes work. |
| orchestrate | Drives one ticket through the whole chain until it's ready for human review. |
| ship | Takes any task, ticket or not, to a PR that an independent reviewer agent has already approved. |
| clean-memory | Reports what agent work costs your machine (worktrees, clusters, containers) and reclaims only what you approve. |
What this listing contains, and what it doesn't
Agensi delivers text only, and each script runs in isolation from a temporary file. These skills rely on helper scripts that work together and need the package files on disk: run locks, GitHub identity checks, PR polling, ticket fetching and capacity preflight. For that reason this listing includes only the instruction text (each skill's SKILL.md, policy references and JSON schemas). The bundled scripts and offline tests are not included. Without them, the agent follows the same rules by hand. Expect less automation and less strict enforcement than the full version.
Full version (recommended for real use): https://github.com/ShavitCohen/team-agent-skills. Clone it and run install.sh to get every script, the offline test suite and the package verifier.
Licence
Open source under the MIT licence (the full text is included as LICENSE).
Where it stops
- It never merges, approves or edits tickets.
- It's built for GitHub. Other trackers aren't supported.
- Manual QA needs local resources (containers or a local cluster) on the machine where the agent runs.
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 any payment.
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
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 17 days ago
- Free with an Agensi account
Requires