AI Agent

Archived Agents

Archive of deprecated agents including the orchestrator. Provides historical context on architectural issues, deprecation reasons, migration paths, and replacements like auto-implement.md coordination workflow.

documentation

Install

npx claudepluginhub akaszubski/autonomous-dev --plugin autonomous-dev

Details

Tool AccessAll tools

RequirementsPower tools

Prompt Preview

This directory contains agents that have been deprecated or replaced. **Archived**: 2025-11-03 **Reason**: Architectural issue - "orchestrator agent" was actually Claude coordinating Claude **Problem**: - When Task tool invoked "orchestrator", it loaded orchestrator.md as Claude's system prompt - But it was still the same Claude instance making decisions - Created logical impossibility: "Claude...

Agent Content

Similar Agents

prompt-manager

all tools

Manages AI prompt library on prompts.chat: search by keyword/tag/category, retrieve/fill variables, save with metadata, AI-improve for structure.

prompts.chat

159.9k

skill-manager

all tools

Manages AI Agent Skills on prompts.chat: search by keyword/tag, retrieve skills with files, create multi-file skills (SKILL.md required), add/update/remove files for Claude Code.

prompts.chat

159.9k

Skill Review: [skill-name]

all tools

Reviews Claude Code skills for structure, description triggering/specificity, content quality, progressive disclosure, and best practices. Provides targeted improvements. Trigger proactively after skill creation/modification.

plugin-dev

83.2k

Stats

Parent Repo Stars10

Parent Repo Forks0

Last CommitDec 24, 2025

Actions

View Source View Plugin View on GitHub View README

Archived Agents

This directory contains agents that have been deprecated or replaced.

orchestrator.md

Archived: 2025-11-03 Reason: Architectural issue - "orchestrator agent" was actually Claude coordinating Claude

Problem:

When Task tool invoked "orchestrator", it loaded orchestrator.md as Claude's system prompt
But it was still the same Claude instance making decisions
Created logical impossibility: "Claude, invoke yourself to tell yourself to invoke agents"
Result: Claude could still make adaptive decisions to skip agents

Solution:

Moved all coordination logic to commands/auto-implement.md
Made Claude's coordination role explicit
No separate "orchestrator agent" - just coordination instructions directly to Claude
Same checkpoints, simpler architecture, more honest design

Migration:

Old: /auto-implement → invokes orchestrator agent → orchestrator coordinates
New: /auto-implement → Claude reads coordination steps → Claude coordinates directly

Replacement: See plugins/autonomous-dev/commands/auto-implement.md for the new coordination workflow

Files affected by this change:

commands/auto-implement.md - Now contains all coordination logic
hooks/enforce_pipeline_complete.py - New enforcement hook
All documentation updated to remove "orchestrator agent" references

This file is kept for historical reference and in case we need to restore parts of the logic.