> **Note:** `CLAUDE.md` is a symlink to this file.

This file defines coordinated multi-agent workflows for enterprise accessibility scanning in the Claude Code environment.

title: "Parallel Agents"

**CRITICAL**: All agents MUST follow the task coordination protocol:

**Version:** 2.0.0 | **Astro 5.x** | **Updated:** 2026-03-13

Workflow automation agents using Gemini CLI, Groq API, and Amp CLI.

55 cross-cutting utilities: path resolution, token truncation, config parsing, model resolution.

Gemini MCP (ACP bridge) routing and delegation rules

Spawn multiple parallel coding agents to work on issues simultaneously.

implement-feature dispatches multiple agents concurrently into the same working directory. Two agents editing the same file simultaneously produce merge conflicts or silent overwrites. The Agent tool supports `isolation: 'worktree'` but implement-feature never uses it.\n\nSymphony enforces three workspace invariants per issue: agent runs only inside workspace_path; workspace_path must be inside workspace_root; workspace key characters restricted to [A-Za-z0-9._-].\n\n**Proposed behaviour:**\n- implement-feature: set `isolation: 'worktree'` on all Agent tool calls for task execution.\n- Worktree naming: derive from task ID using Symphony's sanitization rule (replace non-[A-Za-z0-9._-] with `_`): e.g., `.worktrees/task-1-2/`.\n- After agent completes in the worktree, merge changes back to the main branch (or flag conflicts for operator resolution).\n- start-task skill: detect if running inside a worktree and adapt file path references accordingly.\n\n**Acceptance criteria:**\n- Concurrent task agents each write to a distinct worktree; no shared file edits between agents on separate tasks.\n- Worktree is cleaned up automatically if the agent makes no changes (already the default Agent tool behaviour).\n- Merge-back step produces a clear conflict report if two agents modified the same file.

Show the full AI agent reference guide for this project.

Guide for adding or updating model pricing entries in Langfuse. Use this when

> Best practices for building agents and agentic applications with Plano — the AI-native proxy and dataplane. Covers configuration, routing, agent orchestration, filter chains, observability, CLI oper

Use this guide when Codex should operate a Vault managed by the `LLM Wiki Karpathy` MCP server.

A quick reference for applying 12 Factor Agents principles to multi-agent system design. Based on the [12 Factor Agents](https://github.com/humanlayer/12-factor-agents) framework.

One of the FOCUS research's most profound contributions challenges a deeply held assumption in AI: that expertise requires comprehensive, coherent mental models. The researchers found that "developing

When you have something valuable to teach an AI coding assistant, you face a fundamental choice: package it as a **skill** (a portable SKILL.md file) or build it as an **agent** (a bundle with custom

Blazor is a .NET web framework for building interactive client-side web UIs using C# and Razor syntax. It supports multiple hosting models: Blazor Server (SignalR-based), Blazor WebAssembly (client-si

Modern Python packaging has converged on `pyproject.toml` as the single source of truth for project metadata, build configuration, and tool settings. A series of PEPs have standardized how Python proj

.NET Worker Services are long-running background applications built on the Generic Host (`Microsoft.Extensions.Hosting`). They use `BackgroundService` or `IHostedService` to run tasks that operate ind

|v1.0.0|Velt|March 2026

This document compiles all rules from the `nodejs-solid-architecture` skill into a

> **Note:** This document is a compiled reference for AI agents and LLMs. It

This document contains detailed explanations, code examples, and implementation guidance for all Inertia Rails best practices.

- **intent-gate/** — Intent classification and ambiguity detection. Injected into system prompt.

This document compiles all rules from the Structured JSON Logging Best Practices framework, organized by impact priority for comprehensive structured logging implementation guidance.

This file provides guidelines for agents using the **Plugin Integration Development** skill to assist users in creating, developing, and integrating plugins, tools, and commands into OpenClaw.

Headless browser automation server for AI agents. Run locally or deploy to any cloud provider.

|v1.0.0|Velt|March 2026

Setup guide for using Odoo 18 documentation with AI coding assistants (Cursor, Claude Code, Windsurf, Aider, etc.).

All Meet runtime code lives under this directory. The daemon module, tools,

Run a multimodal Visual Language Model (VLM) with speech interaction on reComputer Jetson (AGX Orin 64G or Orin NX 16G), combining NVIDIA VLM, SenseVoice speech-to-text, and Coqui-ai TTS for voice-driven visual scene understanding.

A comprehensive collection of guidance for Python software development, covering project configuration, dependency management, CLI tooling, and best practices across the modern Python ecosystem.

supabase-postgres-best-practices/

Quick-reference decisions, checklists, and tables for authentication and authorization.

threat-model/

`CommunityToolkit.Diagnostics` provides the `Guard` class, a set of static helper methods for writing concise and performant guard clauses. Guard clauses validate method arguments and invariants at th

- **[OPERATIONAL_RISKS.md](OPERATIONAL_RISKS.md)** - Operational risks, inconsistencies, and incident response procedures

This document is the authoritative reference for AI agents and developers

|v2.0.0|Velt|January 2026

|v1.0.0|Velt|April 2026

You are the Safety department. You handle security audits and guardrails.

Workflow for updating `AGENTS.md` and optionally `DOCS.md` files to match actual codebase state. `CLAUDE.md` is a symlink to `AGENTS.md` and does not need separate processing. DOCS.md is optional and

**Source:** https://claude.com/blog/claude-managed-agents

Deploy Isaac ROS NVBlox with Orbbec RGB-D camera on Jetson AGX Orin for real-time GPU-accelerated 3D mapping (TSDF/ESDF). Sets up Docker CE, Isaac ROS 3.2, Orbbec SDK ROS2, and builds the NVBlox workspace for obstacle detection and 3D mesh generation. Requires JetPack 6.2 and ROS2 Humble.

Deploy a knife detection model on Triton Inference Server using reComputer J1010 for X-ray security scanning, with Raspberry Pi clients sending images for inference and displaying detection results.

This document compiles all rules from the Celery Development Best Practices framework, organized by impact priority for comprehensive distributed task processing guidance.

This document compiles all 50+ rules from the FastAPI Development Best Practices framework, organized by impact priority for comprehensive modern API development guidance.

This document provides guidance for AI agents working with this skill to ensure consistency and avoid common pitfalls.

>-