Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Appearance settings

AppVerk/av-opencode-plugins

Open more actions menu

Repository files navigation

AppVerk OpenCode Plugins

Package

OpenCode plugin packages for AppVerk. The root plugin loads the AppVerk plugin bundle from this repository, which currently provides:

  • A controlled commit workflow (/commit) that enforces AppVerk git policies.
  • A Python development workflow (/python) with TDD, coding standards, and stack-specific patterns (FastAPI, Django, Celery, SQLAlchemy, Pydantic).
  • A TypeScript + React development workflow (/frontend) with TDD, coding standards, and stack-specific patterns (Tailwind, Zustand, TanStack Query, React Hook Form, TanStack Router).
  • A code review workflow (/review) with parallel security and quality audits, verification agents, fix commands, feedback analysis, and skill-agent integration.
  • A QA workflow (/create-qa-plan, /run-qa) for end-to-end testing — generates test plans from PR descriptions and executes them via Playwright (frontend) or HTTP + DB (backend).
  • A Swift development workflow (/swift) with TDD, coding standards, and modern Apple stack patterns (SwiftUI, @Observable, SPM, SwiftData).
  • A global skill registry that makes all AppVerk development skills available to every OpenCode agent via a single load_appverk_skill tool, with mandatory activation rules injected into every agent's system prompt.

Installation

Add the root plugin package to your OpenCode config:

{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["av-opencode-plugins@git+https://github.com/AppVerk/av-opencode-plugins.git#v0.2.16"]
}

Restart OpenCode after updating the config. The root plugin installs the AppVerk plugin bundle and registers /commit, /python, /frontend, /swift, /review, /fix, /fix-report, and /analyze-feedback automatically.

Usage

Commit workflow

Create a commit for the current repository changes:

/commit

Create a commit and append a work item reference:

/commit AV-42

The command uses the packaged AppVerk workflow, generates a Conventional Commit style message, and routes the final commit through the controlled runtime instead of allowing raw git commit from the bash tool.

/python — Python development workflow

Run the Python development workflow with TDD and stack-specific patterns:

/python Add user authentication endpoint with JWT

The /python command:

  1. Detects your project stack (FastAPI, Django, Celery, etc.)
  2. Loads relevant Python development skills
  3. Follows TDD: writes tests first, then implementation
  4. Runs quality gates (typecheck, tests, lint)

You can also invoke the agent directly:

opencode agent python-developer "Refactor user service to use repository pattern"

/frontend — TypeScript + React development workflow

Run the TypeScript + React development workflow with TDD and stack-specific patterns:

/frontend Add user profile form with validation

The /frontend command:

  1. Detects your project stack (Tailwind, Zustand, TanStack Query, etc.)
  2. Loads relevant TypeScript + React development skills
  3. Follows TDD: writes tests first, then implementation
  4. Runs quality gates (typecheck, tests, lint)

You can also invoke the agent directly:

opencode agent frontend-developer "Refactor auth store to use Zustand"

/swift — Swift development workflow

Run the Swift development workflow with TDD and modern Apple stack patterns:

/swift Add user profile screen with SwiftData persistence

The /swift command:

  1. Detects your project stack (SwiftUI, SwiftData, URLSession, etc.)
  2. Loads relevant Swift development skills
  3. Follows TDD: writes tests first, then implementation
  4. Runs quality gates (build, tests)

You can also invoke the agent directly:

opencode agent swift-developer "Refactor networking layer to use async/await"

Global Skill Registry

All AppVerk development skills are automatically available to every OpenCode agent (built-in and custom) through the load_appverk_skill tool. When any agent starts a chat session, mandatory activation rules are injected into its system prompt.

Load a skill manually at any time:

Use the load_appverk_skill tool with name "python-coding-standards"

Available skills include:

  • python-coding-standards, python-tdd-workflow, fastapi-patterns, sqlalchemy-patterns, pydantic-patterns, async-python-patterns, uv-package-manager, django-web-patterns, django-orm-patterns, celery-patterns
  • frontend-coding-standards, frontend-tdd-workflow, tailwind-patterns, zustand-patterns, tanstack-query-patterns, form-patterns, tanstack-router-patterns, pnpm-package-manager
  • standards-discovery (code review)

Code review workflow

Run a comprehensive code review with parallel security and quality audits:

/review Review the authentication module for security issues and code quality

The /review command:

  1. Detects your project stack and loads relevant skills (Python, Frontend, PHP)
  2. Launches security-auditor, code-quality-auditor, and optionally documentation-auditor agents in parallel
  3. Runs verification agents (cross-verifier, challenger) to validate findings
  4. Aggregates findings with unique issue IDs (SEC-001, ARCH-001, etc.)
  5. Generates a structured markdown report
  6. Optionally saves to docs/reviews/YYYY-MM-DD-<branch>.md

Fix commands

Fix a single issue by ID from a saved report:

/fix SEC-001

Or paste the full issue block directly:

/fix [paste issue block from /review or /run-qa]

Batch-fix issues from a saved report:

/fix-report docs/reviews/2026-04-22-feature-auth.md

Feedback analysis

Analyze PR comments and generate response drafts:

/analyze-feedback 123

Classifies each comment as "Address" or "Reject" and optionally publishes responses via GitHub CLI.

QA workflow

Generate a structured test plan from a PR description or ticket:

/create-qa-plan Add two-factor authentication to the login flow

The /create-qa-plan command:

  1. Parses the feature description or ticket text
  2. Generates test cases with preconditions, steps, and expected results
  3. Writes the plan to docs/testing/plans/YYYY-MM-DD-<topic>-test-plan.md

Run a saved test plan or perform an ad-hoc QA check:

/run-qa docs/testing/plans/2026-04-29-feature-auth-test-plan.md

The /run-qa command:

  1. Loads the test plan or creates a quick checklist for the given path
  2. Detects whether the scope is frontend, backend, or both
  3. Launches testing agents in parallel when both scopes exist (@qa-fe-tester and/or @qa-be-tester)
  4. Executes tests using native Playwright tools (frontend) or native HTTP and DB CLI tools (backend)
  5. Collects results into a markdown report with pass/fail status

You can also invoke testing agents directly:

opencode agent qa-fe-tester "Run accessibility checks on the checkout page"
opencode agent qa-be-tester "Test GET /api/v1/orders with pagination"

Available Commands & Agents

Command / Agent Description Mode Docs
/commit Controlled commit workflow — Conventional Commit messages, bash-level blocking for direct git commit/git push. Guide
/python Python development workflow — TDD, coding standards, and stack-specific patterns (FastAPI, Django, Celery, SQLAlchemy). Guide
/frontend TypeScript + React development workflow — TDD, coding standards, and stack-specific patterns (Tailwind, Zustand, TanStack Query). Guide
/swift Swift development workflow — TDD, coding standards, and modern Apple stack patterns (SwiftUI, @Observable, SPM). Guide
/review Code review workflow — parallel security, quality, and documentation audits with verification and structured reports. Guide
/fix Fix a single issue by ID or pasted issue block from a saved review or QA report. Guide
/fix-report Batch-fix issues from a saved review or QA report with interactive selection. Guide
/analyze-feedback Analyze PR feedback comments, classify validity, and generate response drafts. Guide
/create-qa-plan Generate a structured QA test plan from a PR description or ticket. Guide
/run-qa Execute a saved test plan or ad-hoc QA check via Playwright or HTTP + DB. Guide
load_appverk_skill Load any AppVerk development skill by name. Available to all agents globally. Guide
@python-developer Direct agent invocation for Python tasks outside of /python. primary Guide
@frontend-developer Direct agent invocation for TypeScript + React tasks outside of /frontend. primary Guide
@swift-developer Direct agent invocation for Swift tasks outside of /swift. primary Guide
@security-auditor Direct agent invocation for security audits with skill-agent delegation. subagent Guide
@code-quality-auditor Direct agent invocation for code quality audits with skill-agent delegation. subagent Guide
@documentation-auditor Documentation audit agent — verifies code changes are reflected in docs. subagent Guide
@cross-verifier Cross-domain correlation agent — finds intersections between findings. subagent Guide
@challenger Adversarial review agent — challenges findings for false positives. subagent Guide
@synthesis-agent Planned — deduplicates and groups findings into actionable PRs. Not yet implemented. subagent Guide
@qa-fe-tester Frontend testing subagent — runs Playwright tests, accessibility checks, and visual regression. subagent Guide
@qa-be-tester Backend testing subagent — tests API endpoints and validates database state via HTTP + DB CLI. subagent Guide
@feedback-analyzer Per-comment classification agent for PR feedback analysis. subagent Guide
@fix-auto Auto-fix subagent — performs fixes without user confirmation. subagent Guide

Note on Mode: Commands always appear in tab-completion. Agents marked primary also appear in tab-completion, while subagent agents are hidden and must be invoked explicitly (e.g., @fix-auto).

Repository Structure

  • packages/commit - plugin source, tests, command template, and build scripts for the commit workflow.
  • docs/plugins/commit.md - package-level behavior and usage guide.
  • packages/python-developer - plugin source, tests, skill files, and build scripts for the Python development workflow.
  • docs/plugins/python-developer.md - package-level behavior and usage guide.
  • packages/code-review - plugin source, tests, agent prompts, command template, and build scripts for the code review workflow.
  • docs/plugins/code-review.md - package-level behavior and usage guide.
  • packages/frontend-developer - plugin source, tests, skill files, and build scripts for the TypeScript + React development workflow.
  • docs/plugins/frontend-developer.md - package-level behavior and usage guide.
  • packages/skill-registry - global skill registry source, tests, and build scripts.
  • docs/plugins/skill-registry.md - package-level behavior and usage guide.
  • packages/qa - QA plugin source, tests, command templates, agent prompts, skills, and build scripts.
  • docs/plugins/qa.md - package-level behavior and usage guide.
  • packages/swift-developer - plugin source, tests, skill files, and build scripts for the Swift development workflow.
  • docs/plugins/swift-developer.md - package-level behavior and usage guide.
  • package.json - workspace definition and shared validation commands.

Local Development

Install workspace dependencies:

npm install

Run the main validation commands:

npm run typecheck
npm run test
npm run build
npm run check

Documentation

License

This repository currently does not include a top-level LICENSE file. Add one before publishing or distributing the packages beyond internal use.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Morty Proxy This is a proxified and sanitized view of the page, visit original site.