Open Mercato Skills
73 skills for a delivery pipeline: plan, code, review and QA, ending in a ready, labeled PR in your own repository.
- 272 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- Autonomous skills create branches, open PRs, set labels and, depending on setup, carry changes to merge without a human.
- QA skills drive a real browser and run the repository's validation commands.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/open-mercato-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Autonomous skills create branches, open PRs, set labels and, depending on setup, carry changes to merge without a human.
- QA skills drive a real browser and run the repository's validation commands.
This is third-party code. Review the repository files before installing.
What it does
Open Mercato Skills is a set of 73 skills that carry a task from idea to merged pull request. Skills prefixed `om-auto-` run autonomously: given a brief, a spec or an issue, the agent drafts a plan, codes in an isolated worktree, runs the repository's validation commands, reviews its own diff and opens a labeled PR with a summary comment. Skills without `auto` are interactive: they act once, ask questions when needed and hand control back. One group stops before publishing. `om-fix-issue` and `om-implement-feature` leave a validated diff staged plus a worktree path, and `om-setup-agent-harness` adds extra reviewers such as Codex, DeepSeek, Kimi, GLM and MiMo. GitHub is the default tracker, GitLab and Forgejo have their own providers, and Linear and Jira have ready templates.
Who it is for. Teams whose work lives in issues and PRs and who want an agent to carry it to a reviewed, QA-checked PR.
Good fit when
- A bug is described in an issue and the agent should find the cause, fix it and add regression tests.
- A feature is written up as a spec that the agent implements step by step in a separate worktree.
- Open PRs pile up without review and the agent works through them in a batch with verdicts and labels.
- You want to see a staged diff with opinions from several models before anything is published.
- The team process should be written into `SDLC.md` and a config that every skill reads.
Not a fit when
- The repository has no tests or linters: the pipeline relies on validation commands it runs before a PR.
- The agent must not open PRs or change tracker labels without a human.
- The task is small and one-off, and setup through `/om-setup-agent-pipeline` takes longer than the edit itself.
Example request
/om-auto-create-pr "add rate limiting to the login endpoint"Limitations
The pipeline needs a repository with validation commands and rights to branches, PRs and labels in the tracker. Autonomous skills pick the most reversible option themselves and publish the result, so attach them to a repository with a protected main branch. The `om-prepare-test-env`, `om-integration-tests` and `om-auto-qa-pr` skills drive a real browser, so skills.sh validation may flag them as Medium or High. The multi-model mode needs access to the chosen providers. After `npx skills update -p` run `/om-apply-upgrade-notes`, otherwise files created by setup stay outdated.
How to disable. Remove the installed skills with the `skills` manager, and delete the `.ai/` directory and `SDLC.md` if you no longer need them. Close any PRs the agent opened by hand.
FAQ
How do I install Open Mercato Skills in Claude Code?
Run `npx skills add open-mercato/skills --skill '*'` to install all 73 skills. Without `--skill '*'` the installer lets you pick skills interactively.
What should I do right after installing?
Run `/om-setup-agent-pipeline` once per repository. It inspects the repo, asks a few questions, writes `.ai/agentic.config.json` and generates `SDLC.md`, which the other skills read.
How do autonomous skills differ from the rest?
The `om-auto-` prefix means the skill runs without questions from start to PR, claiming the task with an `in-progress` label. Skills without the prefix act once and ask.
Can I stop before anything is published?
Yes. `om-fix-issue` and `om-implement-feature` leave a validated diff staged, a prepared PR body and a worktree path, and you commit and publish yourself.
How do I update the skills?
Run `npx skills update -p` in the project directory, or `npx skills update -g` for a global install. Then run `/om-apply-upgrade-notes` to apply migrations while keeping your edits.
Discussion
No questions or reviews yet.
Related
SkillsFifteen skills from Warp: CI repair, pull requests, accessibility and SEO audits, web app testing and issue triage
Tech Leads Club Agent Skills
Agent Skills
A vetted skill registry with a CLI that installs into twenty-plus agents and an MCP server for on-demand skill lookup
Builder Skills
Skills for agents
A set of small skills from Builder.io for agent work habits: visual plans and diff recaps, docs-first checks, cross-agent review and clean status updates
Nine JS Mastery skills take a change through phases: from idea and spec to verification, tests, release texts and a docs sync
Dimillian Skills
Skills
A set of 16 engineering skills for agents: Apple platform work, macOS packaging, React performance, code review and refactoring
A plugin marketplace and installer for the full dev loop: task selection, review, cleanup, PRs and releases
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Open Mercato Skills (this card) | no | 272 | |
| Superpowers | no | 313,096 | sets the development process inside an agent session; here tasks are carried to a PR with labels, a review loop and a QA gate in the tracker |
| Compound Engineering | no | 25,951 | a brainstorm, plan, work and review cycle that records lessons; here autonomous skills cover issues, specs and merging |
| gstack | no | 140,343 | a virtual team of roles inside Claude Code; here the pipeline is tied to issues, PRs and tracker labels |
README in short
The project grew out of the Open Mercato team's practice: its authors say the ERP product was built through this pipeline with no hand-written code. The repository keeps the process itself, stripped of product specifics. Each skill installs standalone, so the chain can be assembled in parts.