flora131Community·11 Agent Skills Included

atomic

Verifiable coding runs with workflows, subagents, and review gates

Runs long engineering tasks as explicit, checkable execution graphs with parallel stages, human approval gates, and resumable checkpoints. Replaces manual babysitting and re-prompting with built-in verification, repair loops, and evidence-backed results. Includes skills for code research, spec writing, browser testing, code quality checks, stacked PRs, and cross-session coordination.

484|76

↗

npx skills add flora131/atomic --all -g -y

Available:

Tells the agent how this monorepo is structured, which build and test commands to run, and how to treat internal versus user-facing skills when working in the repository.

All Skills in This Repository (11)

Pure Emerald Level Indicators

📦 In Repo

flora131

init

Generate CLAUDE.md and AGENTS.md files by analyzing a codebase for project metadata.

gh-commit

Generate Conventional Commits messages from staged and unstaged Git changes.

testing-anti-patterns

Identify and resolve common testing anti-patterns in software test suites.

research-codebase

Spawn parallel sub-agents to research codebases and generate documents with code references.

prompt-engineer

Create, improve, and optimize AI prompts using Anthropic best practices.

Submit Diff (Sapling + Phabricator)

Submit code commits as Phabricator diffs using Sapling and CLI tools.

sl-commit

Generate Conventional Commits messages for Sapling SCM repositories.

create-spec

Generate technical design documents from existing research directories.

explain-code

Explain code functionality by analyzing structure, logic, and dependencies.

gh-create-pr

Commits unstaged changes, pushes the current branch to remote, and initiates a pull request.

frontend-design

Generate distinctive frontend interfaces with React or Vue code.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install Atomic?▼

Run `npx skills add flora131/atomic --all -g -y` in your terminal to install all skills in this suite globally.

What does Atomic do differently from a normal coding agent?▼

Atomic builds your engineering process as explicit execution graphs with verification checks, review gates, and human approvals, so work is validated with evidence instead of trusting a single quick pass.

Can Atomic resume a run after it stops?▼

Yes. Runs checkpoint as they go, so killing the process or restarting the machine leaves them retained and resumable instead of lost.

How do subagents and intercom work in Atomic?▼

The subagent skill delegates bounded tasks to specialist agents in parallel, and the intercom skill lets separate sessions message each other to split work, ask questions, and share findings.

Can I use Atomic skills without writing code?▼

Yes. You describe workflows in plain English and Atomic writes the runnable graph, while skills like research-codebase and create-spec produce documents from natural language requests.