
Engineering workflow skills for AI coding agents.
Engineering workflow skills for AI coding agents.
·
Engineering workflow skills for AI coding agents. Plan a change, build it, review it, and merge it, with the checks a careful engineer runs at each step.
Snappedly Skills is a collection of Agent Skills: Markdown instructions that your coding agent loads when a task calls for them. They work in Claude Code, Codex, and the other agents that the Skills CLI supports.
Each skill does one job, such as stress-testing a plan, reviewing a diff, or diagnosing an intermittent bug. The skills hand work to each other. They read your team's conventions from files in your repository, so one set of skills fits teams with different hosts, trackers, and review rules.
[!TIP] For details, such as guides, concepts, and a reference for every skill, see the Snappedly Skills docs.
[!IMPORTANT] These skills direct an AI agent that can edit files, run commands, push code, and merge pull requests. Read the disclaimer before you install them.
Install the whole collection:
npx skills add snappedly/skills --skill '*'The main workflow skills use the practice skills in the Background group, such as code-cleanup, code-review, and pr. If you install only some skills, include those too.
In the repository you work on, type /setup-snappedly-skills in your coding agent. Setup asks about your code host, work tracker, and team workflow, and saves your answers in the repository.
Describe the change you want, then run the main workflow:
/grill → /execute → /clean-up → /submitA person then reviews and merges the PR. To skip that review and let the agent merge the PR itself, end with /deploy instead of /submit.
To run a skill, type its name as a command: /grill. Some skills, such as setup-snappedly-skills, execute, clean-up, submit, and deploy, start only when you type their command. Others, such as grill, code-review, and tdd, load automatically when a task calls for them. If you are not sure which skill fits, type /help-snappedly.
Most work goes through four steps.
flowchart LR
grill["grill<br/>(optional)"] --> execute --> cleanup["clean-up"] --> submit --> review["a person reviews<br/>and merges"] --> merged(["merged"])
cleanup -.-> deploy
deploy -.->|skip human review| merged| Skill | What it does |
|---|---|
grill | Interviews you about a plan until every open decision is settled. When the plan is too big for one PR, it proposes a spec with tracer-bullet tickets for execute to carry out as their blockers close. Skip it when the change is clear: describe what you want, let the agent propose a change, and run execute on that proposal. |
execute | Carries out the settled plan, or a tracker issue you name, such as /execute #42. An issue in the needs-triage state is triaged first: execute moves it to ready-for-agent and continues, or stops with the triage recommendation. Name a planning spec to carry out its tickets whose blockers have closed; every open ticket of the spec in the needs-triage state is triaged first too. Substantial independent pieces, such as separate tickets, run in parallel agents. Then it cleans up, reviews the result, and opens a preview. |
clean‑up | The final check. For an independent look, run it in a fresh session, ideally on a different model or provider. It repeats cleanup and review over everything changed on the branch, fixes clear findings, and shows what it changed. Edits stay uncommitted. |
submit | Commits pending changes, pushes, opens or updates the PR, marks it ready for review, and waits for its checks. A person then reviews and merges it: submit never merges. After review feedback, run it again to push fixes and re-request review. |
deploy | Does what submit does, then merges the PR itself when your workflow allows, closes tickets, and verifies production if your workflow asks. It asks you first if commits from elsewhere or resolved conflicts entered the PR, and stops when the merge needs a person. PRs it plans to merge without review get the agent-merge label, so reviewers can filter them out. |
Each skill lives at skills/<group>/<name>/SKILL.md, and the groups below follow the directories under skills/.
| Skill | What it does |
|---|---|
help‑snappedly | Picks the skill or workflow that fits your current situation. |
setup‑snappedly‑skills | Records your repository's code host, work tracker, team workflow, domain docs, and frontend conventions for the other skills. |
| Skill | What it does |
|---|---|
prototype | Builds a throwaway prototype to answer a design question about logic or UI. |
research | Investigates a question against primary sources and captures cited findings. |
The main workflow shows how these skills fit together.
| Skill | What it does |
|---|---|
grill | Interviews you about a plan until every open decision is settled, and splits a plan too big for one PR into a spec with tracer-bullet tickets. |
execute | Carries out the settled plan, a named issue, or a planning spec's ready tickets, with parallel agents where the work splits, then cleans up, reviews, and opens a preview. |
clean‑up | Cleans up, reviews, fixes, and shows what changed, leaving edits uncommitted. |
submit | Opens or updates a PR ready for review, waits for its checks, and leaves the merge to a person. |
deploy | Merges checked changes without human review when your workflow allows, then closes tickets and verifies production. |
| Skill | What it does |
|---|---|
retro | Reviews a coding session and recommends evidence-backed improvements to the agent's environment. |
wait‑what | Explains the last message again, more clearly, when it did not land. |
workflow‑mapping | Maps user journeys and process scenarios, with their outcomes, logic mismatches, and missing cases. |
| Skill | What it does |
|---|---|
codebase‑cleanup | Sweeps a codebase for dead code, useless tests, unnecessary wrappers, and redundant abstractions. |
diagnosing‑bugs | Diagnoses uncertain, persistent, or intermittent bugs and performance regressions. |
improve‑codebase‑architecture | Finds deepening opportunities across a codebase or around a change, and presents them as a visual HTML report. |
triage | Moves issues and external PRs through triage, verifies claims, and writes agent-ready briefs. Agents can load it when needed, including during execute. |
The practice skills hold the standards that the other skills apply. Your agent also loads them when a task calls for them.
| Skill | What it does |
|---|---|
code‑cleanup | Cleans and validates completed changes before a commit. |
code‑review | Reviews a branch, PR, or working change for correctness, requirements, maintainability, and UI, with depth that matches the risk. |
codebase‑design | Supplies a shared vocabulary for designing deep modules and choosing where seams go. |
domain‑modeling | Maintains a project's domain glossary and its architecture decision records. |
pr | Prepares a pull request for review, from its title and description to its commit history. |
remove‑slop | Removes AI-generated slop from code and prose without changing behavior or meaning. |
tdd | Applies test-driven development to changed logic. It also serves as the reference for test quality and test seams. |
technical‑writing | Applies a layered writing standard to docs, PR descriptions, commit messages, and tickets. |
writing‑for‑agents | Guides the writing and editing of skills and AGENTS.md files. |
The skills carry no project-specific policy. setup-snappedly-skills writes your team's choices into the repository, and the other skills read them from there:
docs/agents/code-host.md: your code host, the PR title convention with its approved scopes, and the base branch protection. Setup recommends a rule that blocks direct pushes to the base branch, so every change arrives through a merged PR, which agents can still merge.docs/agents/issue-tracker.md: your work tracker. GitHub Issues is built in. For another tracker, setup records the operations your team uses.docs/agents/workflow.md: the checks to run and which of them two worktrees cannot run at once, the base branch and merge strategy, when change requests open and merge, who merges and who approves the merge, when tickets and specs close, who signs off and who verifies production, and the evidence each step needs.docs/agents/triage-labels.md, docs/agents/domain.md, and docs/agents/frontend.md: triage roles, the layout of your domain docs, and your frontend conventions.The pr skill titles pull requests as type(scope): description by default, with a scope from the approved list. If your repository has its own title convention, that convention wins.
Update your installed skills:
npx skills updateRerun setup-snappedly-skills in each project. Setup keeps your existing conventions and shows each proposed change for your approval.
npx skills update updates only the skills you have installed. To add a skill a release introduced, run npx skills add snappedly/skills --skill <name>. It also does not uninstall a skill that a release removed. It reports the skill as failed and leaves the old copy in place. Remove it with npx skills remove <name>. The changelog lists the skills each release removed.
The Skills changelog lists what changed in each release. CHANGELOG.md has the full details.
Snappedly Skills is provided "as is", without warranty of any kind. You use it entirely at your own risk.
These skills are instructions for AI coding agents. Depending on how you configure and run your agent, the skills can direct it to edit and delete files, run commands, commit and push code, merge change requests, close tickets, and call external services. AI agents make mistakes, and their output can be wrong, incomplete, or insecure.
You are solely responsible for:
Contributions are welcome. Ask questions and share ideas in Discussions. Read CONTRIBUTING.md before you open a pull request. By submitting a contribution, you agree to license it under the MIT License. To report a security issue, follow SECURITY.md.
Skills in this collection are adapted from Matt Pocock's skills. The remove-slop, technical-writing, and pr skills adapt skills from the pstack and cursor-team-kit Cursor plugins. The remove-slop skill also includes writing rules from Humanizer. The code-review skill applies a pinned copy of the Vercel Web Interface Guidelines. All of these projects are released under the MIT License. THIRD_PARTY_NOTICES.md lists the skills that each project covers, with its copyright and license notices. Each skill folder also carries a LICENSE.txt with the notices that apply to it, so installed copies keep them.
Snappedly Skills is released under the MIT License, which contains the full warranty disclaimer and limitation of liability. You may use, copy, modify, and distribute it, including for commercial purposes, if you include the copyright notice and license text in all copies or substantial portions.
Paste this prompt into your agent. It reads this page and does the setup for you.
Read https://aiagentslisting.com/mcp/snappedly-skills to learn what the "Snappedly Skills" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.Agents can also browse this directory over MCP at https://aiagentslisting.com/api/mcp
npx skills add snappedly/skills --skill '*'
npx skills updateThis server runs locally, so we can't read its tool list over the web yet.
We could not run this server: it ships no npx package for our sandbox and no public endpoint. The four checks that need a live server are left out, and the grade tops out at B until one exists.
Paste this line near the top of your repository's README. The badge always renders the current score.
[](https://aiagentslisting.com/mcp/snappedly-skills)Nothing comparable is listed yet.
v1.6.0v1.6.0 · Build Local removedRelease notesv1.5.6v1.5.6 · Reliable tracker reads and fresh subagentsRelease notesv1.5.5v1.5.5 · Quick approval for retro recommendationsRelease notesv1.5.4v1.5.4 · Triage before executeRelease notesv1.5.3v1.5.3 · Clear status for follow-up reportsRelease notesv1.5.2v1.5.2 · Decisions and gaps before you deployRelease notesv1.5.1v1.5.1 · Execute dependency provenanceRelease notesv1.5.0v1.5.0 · Specs and tickets for agentsRelease notesv1.4.5v1.4.5 · Isolated checks and resumable validationRelease notesv1.4.4v1.4.4 · Concurrent check runsRelease notesv1.4.3v1.4.3 · Diagnosing Bugs evidence handlingRelease notesv1.4.2v1.4.2 · Execute authority boundsRelease notesv1.4.1v1.4.1 · Planning specs and tracer-bullet ticketsRelease notesv1.4.0v1.4.0 · Resolving Merge Conflicts removedRelease notesv1.3.0v1.3.0 · Submit and Deploy skillsRelease notesv1.2.1v1.2.1 · Branch protection and CLAUDE.md retirementRelease notesv1.2.0v1.2.0 · Merge approval and close-outRelease notesv1.1.0v1.1.0 · Deploy owns the commitRelease notesv1.0.0v1.0.0 · Public releaseRelease notesConnect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "snappedly-skills" with a rating from 1 to 5 and a short note on what worked and what did not.No reviews yet
Be the first to share your experience with this tool.
One email a week. New agents, MCP servers and skills, and what is actually getting traction.