Reading
Devflow is a development workflow plugin specifically designed for Claude Code, offering 26 automation skills, 27 professional agents, and 25 lifecycle hooks. It supports the full development process from requirement analysis to code merging. The plugin enables parallel reviews via Agent Teams, supports automatic PR reviews, debugging workflows, and Jira integration, significantly boosting development efficiency.
Section 01
Devflow is a Claude Code plugin designed to automate the full software development workflow from requirement analysis to code merge. It features 26 automation skills, 27 professional agents, and 25 lifecycle hooks, leveraging Agent Teams for parallel PR reviews, supporting Jira integration, and significantly boosting development efficiency. This thread breaks down its design, features, installation, and practical use cases.
Section 02
Against the backdrop of AI-assisted programming's growth, integrating AI into development workflows remains a challenge. Devflow positions itself as a structured plugin for Claude Code (unrelated to Foundry's devflow) focused on general development automation. Its core goals include standardizing workflows to reduce repetitive decisions, enabling parallel task processing via Agent Teams, providing reusable skills and agent configurations, and covering the full development lifecycle. Key components: 26 auto-triggerable skills, 27 specialized agents, 25 lifecycle hooks (dependency checks, quality gates), 4 output styles (senior engineer, coding mentor, Thai/English variants), and a TypeScript SDK for programmatic PR reviews and security audits.
Section 03
Pre-requisites: Install mandatory tools like jq, authenticated GitHub CLI (gh), and git. For macOS users, use Homebrew to install these. Recommended tools: rtk (reduce token usage), shellcheck (validate .sh files), node + markdownlint-cli2 (lint .md files), fd (fast file search), ast-grep (precise code search). Plugin installation: Add the marketplace (wasikarn/devflow) → install the plugin → enable Agent Teams via claude config set env.CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS 1. Restart Claude Code to check for missing dependencies.
Section 04
Devflow offers 8 key skills covering daily development scenarios:
/devflow:work-context: Gather sprint tickets, open PRs, and unmerged branches./devflow:build: End-to-end development from Jira ticket or natural language to PR creation./devflow:review: Parallel PR review (3 agents) covering correctness, architecture, security, and test/documentation completeness./devflow:respond: Auto-fix PR comments, submit changes, and reply to reviewers./devflow:debug: Analyze errors and generate fix PRs./devflow:merge-pr: Perform pre-checks, merge PRs, and manage tags.Typical workflow: Start → work-context → decision branch (new task→build→PR; PR review→respond; bug→debug) → merge PR. Jira integration: Connect Jira instances, map project keys, manage ticket states, and create branches/PRs directly from tickets.
Section 05
Agent Teams: Mandatory for parallel processing (e.g., 3 agents reviewing PRs simultaneously). Disabling this feature downgrades to single-agent mode. Output styles: 4 options (senior engineer: concise technical feedback; coding mentor: educational content; Thai/English variants for localization). SDK: TypeScript devflow-engine SDK for programmatic access (e.g., PR review, security audit) to integrate into CI/CD pipelines.
Section 06
Recommended ecosystem: Git + GitHub CLI (version control), Homebrew/apt (package management), shellcheck/markdownlint (code quality), fd/ast-grep (search tools), rtk/zsh (terminal enhancement). Troubleshooting: Fix marketplace add failures with HTTPS URLs, verify Agent Teams env var, check dependencies via claude plugin list, re-authenticate gh if needed. Best practices: Create issues before major work, keep the plugin updated, familiarize with built-in skills before customization, and version-control workflow configurations.
Section 07
Devflow represents the evolution of AI-assisted development tools—from single code completion to full workflow orchestration, and from passive response to active agent collaboration. It benefits individuals (reduces repetitive work), teams (standardizes workflows and improves code quality consistency), and organizations (supports compliance via auditable processes). As Claude Code and Agent Teams evolve, Devflow will play an increasingly important role in turning AI into a collaborative partner in software development.
Keep going with more reads from the same topic.
Splinter is a minimalist, high-performance key-value (KV) and vector storage system enabling zero-latency inter-process communication via shared memory and atomic operations. With only 766 lines of core code, it supports millions of operations per second and 768-dimensional vector storage, offering a new architectural approach for local LLM inference and data-intensive applications.
LiteMind is a Python framework that provides developers with a unified API to integrate mainstream LLM providers such as OpenAI, Anthropic, Google Gemini, and Ollama. It supports multimodal input/output, tool calling, RAG enhancement, and agent construction.
OmniRoute is an open-source AI gateway that integrates over 67 large model providers through a single OpenAI-compatible endpoint. It offers intelligent routing, load balancing, automatic degradation, and quota management, helping developers get the most available AI services at the lowest cost.
This article introduces an open-source multimodal RAG framework based on Google Gemini Embedding 2, which can uniformly handle embedding and retrieval of four media types: text, images, videos, and audio. Combined with Supabase pgvector vector database and OpenRouter large language models, it provides a complete production-level retrieval-augmented generation pipeline.