PHPackages rhazen35/agent-orchestration - PHPackages - PHPackages [Skip to content](#main-content)[PHPackages](/)[Directory](/)[Categories](/categories)[Trending](/trending)[Leaderboard](/leaderboard)[Changelog](/changelog)[Analyze](/analyze)[Collections](/collections)[Log in](/login)[Sign up](/register) 1. [Directory](/) 2. / 3. [Utility & Helpers](/categories/utility) 4. / 5. rhazen35/agent-orchestration ActiveLibrary[Utility & Helpers](/categories/utility) rhazen35/agent-orchestration ============================ A tool-agnostic agent orchestration setup for spec-driven development. Provides skills, steering, templates, and a factory pipeline workflow. v0.0.3(3mo ago)03MIT Since Mar 26Pushed 2mo agoCompare [ Source](https://github.com/rhazen35/agent-orchestration)[ Packagist](https://packagist.org/packages/rhazen35/agent-orchestration)[ Docs](https://github.com/rhazen35/agent-orchestration)[ RSS](/packages/rhazen35-agent-orchestration/feed)WikiDiscussions main Synced 3w ago READMEChangelog (3)Dependencies (1)Versions (5)Used By (0) [![Agent Orchestration](assets/header.png)](assets/header.png) rhazen35/agent-orchestration ============================ [](#rhazen35agent-orchestration) **A tool-agnostic agent orchestration setup for spec-driven development.** Give your AI agent a structured, auditable factory pipeline — from vague idea to shipped code — without burning your token budget. [![License: MIT](https://camo.githubusercontent.com/08cef40a9105b6526ca22088bc514fbfdbc9aac1ddbf8d4e6c750e3a88a44dca/68747470733a2f2f696d672e736869656c64732e696f2f62616467652f4c6963656e73652d4d49542d626c75652e737667)](LICENSE)[![Packagist](https://camo.githubusercontent.com/150edffe428392b6d2df43174c9498e1077f88da8ab013e062127a99a8d50daf/68747470733a2f2f696d672e736869656c64732e696f2f7061636b61676973742f762f7268617a656e33352f6167656e742d6f726368657374726174696f6e)](https://packagist.org/packages/rhazen35/agent-orchestration)[![npm](https://camo.githubusercontent.com/72b3e0a022ff90bf886941aeb14e0841d160126509b956010cb3c0fac6eb57cb/68747470733a2f2f696d672e736869656c64732e696f2f6e706d2f762f7268617a656e33352d6167656e742d6f726368657374726174696f6e)](https://www.npmjs.com/package/rhazen35-agent-orchestration) --- What is this? ------------- [](#what-is-this) This package provides a complete **Factory Pipeline** workflow for AI agents. It includes skills, steering rules, templates, and governance guardrails — all as plain markdown files that any agent can read, regardless of language or tooling. No runtime code. No dependencies. Works with PHP, Go, Rust, Python, Node, or anything else. --- The Factory Pipeline -------------------- [](#the-factory-pipeline) > Every phase is **manually triggered**. Nothing runs automatically. You stay in control, and your token budget stays intact. ``` Design → Queue → Production → Sentinel → Quality Control → Archiving ``` ### 🧠 Phase 1 — Design [](#-phase-1--design) **Trigger:** "I want to build X" or "Start a new spec for Y" The `spec-architect` skill orchestrates sub-skills in sequence: Sub-skillWhat it does`discovery`Maps the codebase, checks cache before re-scanning`mcp-orchestrator`Discovers available MCP tools, enriches spec with live data`grill-me`Challenges assumptions, provides recommended answers`acceptance-criteria-generator`Drafts requirements and testable acceptance criteria**Output:** signed-off `plan/{namespace}/plan.md`**Gate:** you must approve the plan before anything enters the inbox. --- ### 📥 Phase 2 — Queue [](#-phase-2--queue) **Trigger:** "queue tasks" or "load inbox" Tasks are loaded into `inbox/{namespace}/tasks.md` in **batches of max 3** (the Hydration Limit). The next batch only loads after a slot opens. A flawed plan fails fast and cheap, not after 20 API calls. **Gate:** confirm the inbox before production starts. --- ### ⚙️ Phase 3 — Production [](#️-phase-3--production) **Trigger:** "process next task" The `task-processor` runs with **JIT context** — it only reads the task, the file it's editing, and `steering/principles.md`. Nothing else. Small context window, fast and cheap calls. **Output:** task moved to `outbox/{namespace}/pending-review.md` --- ### 🛡️ Phase 3.5 — Sentinel Pre-Check [](#️-phase-35--sentinel-pre-check) **Trigger:** runs as part of "review", before the quality reviewer The `sentinel` catches cheap failures before they reach the expensive inspector: - Syntax errors or lint failures - Ghost comments (`TODO`, `FIXME`, unfinished notes) - Changelog format errors - Scope creep (files touched outside the task's scope) **Fail:** task returns to `inbox/` with `[SENTINEL_FAIL]`. Quality reviewer is never called. --- ### 🔍 Phase 4 — Quality Control [](#-phase-4--quality-control) **Trigger:** "review" The `quality-reviewer` performs deep reasoning against `acceptance-criteria.md` and `steering/principles.md`. Also runs `integrity-check` for ghost features and architectural drift. ResultAction✅ PassTask moves to `outbox/{namespace}/completed.md`❌ FailTask returns to `inbox/` with `[REJECTION NOTE]` + incremented `[LOOP COUNT]`> If a task is rejected **4 times**, the pipeline locks. Human intervention required. --- ### 📦 Phase 5 — Archiving [](#-phase-5--archiving) **Trigger:** "archive" The `changelog-architect` skill: 1. Appends an entry to `CHANGELOG.md` under `## [Unreleased]` 2. Appends the task + review notes to `archive/{namespace}/decision-log.md` 3. Cleans up `outbox/{namespace}/completed.md` 4. Invokes `monitor` to update `dashboard.md` --- Quick Reference --------------- [](#quick-reference) CommandPhase triggered"I want to build X"🧠 Phase 1: Design"Queue tasks"📥 Phase 2: Queue"Process next task"⚙️ Phase 3: Production"Review"🛡️ + 🔍 Sentinel → Quality Control"Archive"📦 Phase 5: Archiving--- Installation ------------ [](#installation) This package contains no runtime code — only markdown files consumed by your AI agent. It works with any language or stack. ### PHP — Composer [](#php--composer) ``` composer require rhazen35/agent-orchestration ``` Copy the directories into your project root: ``` cp -r vendor/rhazen35/agent-orchestration/skills ./skills cp -r vendor/rhazen35/agent-orchestration/steering ./steering cp -r vendor/rhazen35/agent-orchestration/templates ./templates cp -r vendor/rhazen35/agent-orchestration/records ./records ``` ### Node — npm [](#node--npm) ``` npm install rhazen35-agent-orchestration ``` ``` cp -r node_modules/rhazen35-agent-orchestration/skills ./skills cp -r node_modules/rhazen35-agent-orchestration/steering ./steering cp -r node_modules/rhazen35-agent-orchestration/templates ./templates cp -r node_modules/rhazen35-agent-orchestration/records ./records ``` ### Go, Rust, Python, or any other language — Git Submodule [](#go-rust-python-or-any-other-language--git-submodule) ``` git submodule add https://github.com/rhazen35/agent-orchestration .agent ``` Point your agent tooling at `.agent/steering/`, `.agent/skills/`, etc. Update to the latest version: ``` git submodule update --remote .agent ``` Clone a project that uses this as a submodule: ``` git clone --recurse-submodules ``` ### Quick Start (no package manager) [](#quick-start-no-package-manager) ``` git clone https://github.com/rhazen35/agent-orchestration .agent ``` --- Namespacing ----------- [](#namespacing) Every specification gets its own namespace folder — a short kebab-case name applied consistently across all directories. DirectoryNamespaced pathSpecification`specification/{namespace}/`Plan`plan/{namespace}/`Inbox`inbox/{namespace}/tasks.md`Outbox`outbox/{namespace}/`Records`records/{namespace}/`Archive`archive/{namespace}/`--- Governance ---------- [](#governance) `steering/governance.md` defines hard limits no agent can override: RuleLimitInbox hydrationMax 3 active tasks at onceLoop limitTask rejected 4 times → pipeline locksToken budget50,000 tokens per run → Soft HaltCooldown5 consecutive completions → wait for "Continue"Discovery cacheNever re-scan if fingerprint is currentJIT contextWorker reads only what it needs--- Dashboard --------- [](#dashboard) `dashboard.md` is the control tower. Check it before starting any work. StatusMeaning🟢 OPERATIONALAll clear🟡 THROTTLEDWait for human `[CONTINUE]`🔴 CIRCUIT BREAKER TRIPPEDDo not start work — human reset required--- Skills Reference ---------------- [](#skills-reference) SkillPurpose`spec-architect`End-to-end spec creation`discovery`Codebase mapping with caching`mcp-orchestrator`Dynamic MCP tool discovery and data enrichment`grill-me`Assumption stress-testing`acceptance-criteria-generator`Requirements → testable criteria`task-processor`JIT task implementation`sentinel`Cheap pre-check before review`quality-reviewer`Deep spec + principle review`integrity-check`Drift and ghost feature detection`changelog-architect`Task archiving and changelog`monitor`Dashboard updates--- Steering Files -------------- [](#steering-files) Always loaded into every agent context: FilePurpose`steering/principles.md`Core architectural constraints`steering/governance.md`Hard resource and loop limits`steering/audit-requirements.md`Requirement changes must be logged and ADR'd`steering/workflow.md`Full pipeline reference--- MCP Integration --------------- [](#mcp-integration) This package is designed to work with [Model Context Protocol (MCP)](https://modelcontextprotocol.io) servers out of the box. MCP tools extend the agent's capabilities beyond local file editing — enabling live database inspection, web search, API fetching, and more. ### How it works [](#how-it-works) The agent does not hardcode any MCP servers. Instead, it performs a **handshake** at session start: 1. Agent calls `list_tools` to discover what the host environment provides. 2. Agent matches the returned tool schemas against `steering/mcp-policy.md`. 3. Authorized tools are used purposefully — local files always take precedence. ### Capability Requirements [](#capability-requirements) The factory expects the host to provide tools for these capabilities: CapabilityUsed forFallbackSearchVerify library versions, docsAsk humanFilesystemRead/write specs and inboxLocal agent toolsRuntimeRun tests and linting (Sentinel)Manual checksDatabaseInspect live schemasUse spec definitionsVersion ControlRead issues, PRs, commit historyUse `records/` and `archive/`If a capability is missing, the pipeline continues — it never blocks on absent tools. ### Tool Selection Priority [](#tool-selection-priority) When multiple tools fulfill the same capability, the agent picks in this order: 1. Local filesystem tools (lowest cost) 2. Specialized local servers (e.g., local Postgres, SQLite) 3. Remote API servers (e.g., GitHub, Brave Search, Slack) ### The Local-First Rule [](#the-local-first-rule) Before calling any MCP tool, the agent must check: 1. `specification/{namespace}/` — does the spec already answer this? 2. `archive/mcp-data/{namespace}/` — has this tool been called and cached? 3. `cache/discovery-report.md` — is this covered by the last discovery run? Only if all three are empty or stale does the agent proceed with a tool call. Tool outputs are persisted to `archive/mcp-data/` so the worker never re-calls a tool for the same data. ### MCP in the Pipeline [](#mcp-in-the-pipeline) PhaseMCP usagePhase 1: Design`mcp-orchestrator` runs full capability mapping, enriches spec with live dataPhase 3: Production`task-processor` invokes `mcp-orchestrator` in lightweight mode for specific data needsPhase 3.5: SentinelUses runtime tools (linter, test runner) if available### Relevant Files [](#relevant-files) FilePurpose`steering/mcp-policy.md`Capability requirements, tool selection logic, authorization rules`steering/mcp-governance.md`Hard limits: context budget, write safety, failure fallback`steering/mcp-etiquette.md`Local-First rule, purposeful tool use, no spec overwrites`skills/mcp-orchestrator.md`The bridge skill — discovery, justification, execution, sentinel filter`archive/mcp-data/`Persisted tool outputs, reused across tasks--- Customizing for Your Project ---------------------------- [](#customizing-for-your-project) Every skill, steering file, and template can be overridden at the project level without touching the package files. The agent checks `.agent/` first before falling back to the package defaults. ``` your-project/ .agent/ skills/ quality-reviewer.md ← overrides the default sentinel.md ← overrides the default steering/ principles.md ← overrides the default templates/ plan/ plan-template.md ← overrides the default ``` To extend rather than replace, start your override file with: ``` # [Skill Name] — Project Extension > Extends the base skill. All base behavior applies unless explicitly overridden below. ## Project-Specific Additions ...your additions here... ``` Two files should never be overridden: `steering/workflow.md` (the pipeline contract) and `steering/overrides.md` (the override system itself). See `steering/overrides.md` for the full resolution order, a complete list of overridable files, and guidance on what to override with caution. --- License ------- [](#license) MIT — see [LICENSE](LICENSE). ### Health Score 31 — LowBetter than 66% of packages Maintenance86 Actively maintained with recent releases Popularity3 Limited adoption so far Community6 Small or concentrated contributor base Maturity27 Early-stage or recently created project Bus Factor1 Top contributor holds 100% of commits — single point of failure How is this calculated?**Maintenance (25%)** — Last commit recency, latest release date, and issue-to-star ratio. Uses a 2-year decay window. **Popularity (30%)** — Total and monthly downloads, GitHub stars, and forks. Logarithmic scaling prevents top-heavy scores. **Community (15%)** — Contributors, dependents, forks, watchers, and maintainers. Measures real ecosystem engagement. **Maturity (30%)** — Project age, version count, PHP version support, and release stability. ### Release Activity Cadence Every ~0 days Total 3 Last Release 91d ago ### Community Maintainers ![](https://www.gravatar.com/avatar/cb450f546f3831bdccecd8fde5793abdb503686dfb0b192cdfe3b268940ec9bf?d=identicon)[rhazen35](/maintainers/rhazen35) --- Top Contributors [![rhazen35](https://avatars.githubusercontent.com/u/11785626?v=4)](https://github.com/rhazen35 "rhazen35 (1 commits)") --- Tags aiworkflowAgentorchestrationspec-driven ### Embed Badge ![Health badge](/badges/rhazen35-agent-orchestration/health.svg) ``` [![Health](https://phpackages.com/badges/rhazen35-agent-orchestration/health.svg)](https://phpackages.com/packages/rhazen35-agent-orchestration) ``` ### Alternatives [helsingborg-stad/municipio A bootstrap theme for creating municipality sites. 4028.3k10](/packages/helsingborg-stad-municipio)[maestroerror/laragent Power of AI Agents in your Laravel project 639142.5k](/packages/maestroerror-laragent)[symfony/ai-agent PHP library for building agentic applications. 31746.8k82](/packages/symfony-ai-agent)[mohamed-ashraf-elsaed/claude-agent-sdk-laravel Anthropic Claude Agent SDK for PHP & Laravel — build AI agents with tool use, sandboxing, MCP servers, subagents, hooks, and structured output via the Claude Code CLI 171.1k](/packages/mohamed-ashraf-elsaed-claude-agent-sdk-laravel) PHPackages © 2026 [Directory](/)[Categories](/categories)[Trending](/trending)[Changelog](/changelog)[Analyze](/analyze)