By vitorandtxr
Uma agência de software que busca alavancar e reduzir o tempo de prototipagem e validação técnica de aplicações
Agentic web scraper that reads API documentation from a URL, extracts endpoints, authentication, models, and error handling into a structured corpus, then invokes /skill-creator to generate a new skill that serves as an interactive reference for that API. Use when: (1) user provides a URL to an API documentation site, (2) user says "scrape api docs", "create api reference skill", "import api documentation", "api doc scraper", (3) user wants to create a skill from external API documentation for future technical Q&A.
Progressive Disclosure Optimizer — Analyze and restructure Claude Code configuration files (CLAUDE.md, skills, commands, agents) to minimize token consumption. Supports both project roots (CLAUDE.md) and skill directories (SKILL.md). Applies hierarchical fragmentation, content indexing, skill extraction, and script extraction to reduce startup/trigger context cost. Use when: (1) analyzing repository token budget (/apply-progressive-disclosure analyze), (2) optimizing CLAUDE.md or SKILL.md for token efficiency (/apply-progressive-disclosure optimize), (3) generating before/after token reports (/apply-progressive-disclosure report), (4) optimizing a skill directory (/apply-progressive-disclosure analyze <skill-path>), or (5) user says "optimize context", "reduce tokens", "progressive disclosure", "context bloat", "token budget", "fragment CLAUDE.md", "optimize skill".
Backlog management system — Manages user stories as structured JSON with role-based access control, replacing markdown-based BACKLOG.md. Provides CRUD operations for user stories, status tracking, acceptance criteria, dependency maps, and auto-generated markdown summaries. Enforces agency permissions: only po/pm create stories, only po/pm/tl edit descriptions, all agents can change status. Supports migration from legacy markdown BACKLOG.md via review teams. Use when: (1) any agent needs to create, read, update, or delete user stories, (2) agent mentions "backlog", "user story", "US-", "acceptance criteria", "story status", (3) po/pm is producing backlog in Plan phase, (4) any agent needs to query stories by status/feature/priority, (5) generating BACKLOG.md summary from JSON data, (6) migrating an old docs/BACKLOG.md to the new JSON format.
Developer Specialist — Execution-focused engineer. Writes production code following the approved architecture, conventions from CLAUDE.md, and acceptance criteria from the backlog. Use when: (1) identifying implementation risks during planning (/dev plan), (2) reviewing architecture for implementability (/dev design), (3) writing production code (/dev implement), (4) adding inline code documentation (/dev document), or (5) user says "dev", "developer", "implement", "write code", "build feature", "fix code", "production code".
Project Setup for Omni-SW Agency — Configures a project directory with all base files and hooks required by the SDLC Agency plugin before the orchestrator can run. Creates CLAUDE.md, agent_docs/ directories, notification hooks, and validates the environment. Use when: (1) setting up a new project for the agency (/init-omni-sw), (2) user says "setup", "configure project", "init project", "prepare project for agency", "install hooks", "configure hooks", "setup omni-sw", "configure plugin", or (3) before running the orchestrator for the first time on a project. This skill should also trigger when the user wants to verify or fix the agency configuration in an existing project.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Claude Code plugin that implements a multi-agent SDLC workflow. Five role-based agents collaborate across seven sequential phases to take a project from business objective to delivered, documented software.
Load the plugin when starting a Claude Code session from your project directory:
claude --plugin-dir /path/to/omni-sw
All skills become available namespaced as /omni-sw:<skill>.
Each agent is a skill invoked with a phase argument. You drive the workflow by calling agents in order — each reads the artifacts produced by the previous phase and writes the next one.
/omni-sw:pm plan → /omni-sw:po plan → /omni-sw:tl design
→ /omni-sw:pm validate → /omni-sw:tl validate
→ /omni-sw:dev implement → /omni-sw:tl review
→ /omni-sw:qa test → /omni-sw:pm document → /omni-sw:tl document
Alternatively, let the Orchestrator run the entire workflow automatically:
/omni-sw:orchestrator "Build a REST API for order management"
| Skill | Role | Leads phases |
|---|---|---|
/omni-sw:pm | Product Manager | Plan (biz), Validate (biz), Document (biz) |
/omni-sw:po | Product Owner | — (assists PM, produces backlog) |
/omni-sw:tl | Tech Lead | Design, Validate (tech), Review, Document (tech) |
/omni-sw:dev | Developer Specialist | Implement |
/omni-sw:qa | QA Specialist | Test |
graph LR
Plan --> Design --> Validate --> Implement --> Review --> Test --> Document
Validate -->|"PM reproves"| Plan
Validate -->|"TL reproves"| Design
Review -->|"blocking issues"| Implement
Test -->|"bug failures"| Implement
| # | Phase | Leader | Gate? | Key artifact |
|---|---|---|---|---|
| 1 | Plan | PM → PO | — | {DOCS_PATH}/PROJECT_BRIEF.md, backlog |
| 2 | Design | TL | — | {DOCS_PATH}/ARCHITECTURE.md |
| 3 | Validate | PM + TL | Dual approval | {DOCS_PATH}/VALIDATION.md |
| 4 | Implement | Dev | — | src/ |
| 5 | Review | TL | Blocking issues | {DOCS_PATH}/REVIEW.md |
| 6 | Test | QA | All tests pass | tests/, {DOCS_PATH}/TEST_REPORT.md |
| 7 | Document | PM + TL | — | README.md, CHANGELOG.md, {DOCS_PATH}/API_REFERENCE.md |
{DOCS_PATH} is a timestamped subfolder under agent_docs/ (e.g., agent_docs/build-auth-api-2026_03_21_10_30/), created per SDLC run to isolate working artifacts. The docs/ directory holds maintained descriptive documentation (architecture overview, flow diagrams) for human consumption.
Gate phases enforce quality before progression: both PM and TL must APPROVE to proceed from Validate; blocking issues must be resolved before leaving Review; all tests must pass before reaching Document.
agency_cli.py for deterministic operations)The fastest way is the /omni-sw:init-omni-sw skill, which handles the full configuration in one step:
Start a Claude Code session with the plugin loaded from your project directory:
claude --plugin-dir /path/to/omni-sw
Run the init skill:
/omni-sw:init-omni-sw
The skill detects whether the project directory is empty, has existing code, or already has a CLAUDE.md, and handles each case:
CLAUDE.md/omni-sw:repo-map to discover the stack, then creates CLAUDE.mdBegin the workflow:
/omni-sw:orchestrator "<objective>"
If you prefer to configure manually, copy and fill in the CLAUDE.md template:
cp /path/to/omni-sw/docs/templates/CLAUDE_TEMPLATE.md ./CLAUDE.md
Fill in: Stack, Conventions, Domain Glossary, Forbidden Patterns, Agent Overrides.
Then run setup to create directories and install hooks:
python /path/to/omni-sw/skills/shared/scripts/agency_cli.py setup --scan-root .
The agency is stack-agnostic. Agents read CLAUDE.md and adapt their output accordingly.
The /omni-sw:orchestrator skill drives the full 7-phase workflow end-to-end, spawning agents in the correct sequence, evaluating gates, handling feedback loops, and surfacing questions to you when needed.
For any non-trivial project, write the objective in a dedicated file. Mention it with @ in the chat so Claude Code loads the file content into context before the orchestrator starts — this gives every spawned agent the full business context from the beginning.
Create docs/OBJECTIVE.md in your project root:
## Goal
Build an order management REST API with JWT authentication.
npx claudepluginhub vitorandtxr/omni-sw --plugin omni-swPlugin de ingeniería de software completa: 10 agentes de núcleo y 9 opcionales con personalidad propia, memoria persistente por proyecto, quality gates y flujos automatizados desde la idea hasta producción.
Complete SDLC framework with 58 specialized agents for software development lifecycle management. Phase-based workflows (Inception→Elaboration→Construction→Transition), security reviews, testing orchestration, and deployment automation.
An agent-routed harness for end-to-end software product development
Give soul to your workflow. 58 AI-powered skills across 17 roles — PM, Dev, Backend, Frontend, QA, UX, Data, Detect, WordPress, Release, Security, DevOps, and Core. Spec-to-ship pipeline: scaffold, implement, test, secure, deploy. Features two-phase workflow with human approval, quality-reviewer agent, token optimization, and continuous improvement via LEARN.md system.
Helder's personal SDLC toolbelt for AI coding agents — from PRD to ship. Bundles the tracer-bullet workflow alongside TDD, code review, audits, and shipping skills.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.