updated

Author: utkarshx

docs/.obsidian/workspace.json

@@ -168,8 +168,14 @@
       "command-palette:Open command palette": false
     }
   },
-  "active": "c883305ca140b86c",
+  "active": "9c82afebc22e5868",
   "lastOpenFiles": [
+    "MultiAgentOrchestration/ParallelAgentExecution",
+    "MultiAgentOrchestration/AgentObservabiltiy",
+    "MultiAgentOrchestration/TaskOrchestor",
+    "MultiAgentOrchestration/RemoteTaskExecution",
+    "MultiAgentOrchestration/RemoteAgentExecution",
+    "MultiAgentOrchestration/Tasks",
     "CustomAgents/core/multi-agents/overview.md",
     "MultiAgentOrchestration/index.md",
     "MultiAgentOrchestration",
@@ -199,12 +205,6 @@
     "api/ApiAccess/fs/createFile.md",
     "Docs/portal/8_setting/apikey.md",
     "Docs/application/1_getstarted/index.md",
-    "Docs/application/1_getstarted/1_codeboltide.md",
-    "CustomAgents/Architecture",
-    "CustomAgents/CoreConcepts",
-    "develop2/core/typescript-sdk",
-    "develop2/core/cli",
-    "develop2/core/context",
-    "develop2/core/task-flow"
+    "CustomAgents/Architecture"
   ]
 }

docs/MultiAgentOrchestration/index.md

@@ -2,4 +2,40 @@ Codebolt is an AI Editor build from ground up for Multi-Agent Orchestration. Cod
 
 Codebolt is built for the world where agents are not working for 2-5 min for the task which the user has asked, instead it is built for the world where for a task there are 50-100 agents running, each with different roles, communicating with each other, and building tools and applications, reviewing generated code and even creating custom agents and custom tools on the fly, and working for hours or days to build a complex application. 
 
-Building and Running Multi-Agent Systems have their own problems 
+Building and Running Multi-Agent Systems have their own sets of challenges and need a different set of tools to manage the Multi-Agent Swarm. A lot of things like Agent Observabilty, 
+Task Management, Task Scheduling, LifeCycle Orchestrations Come into Play. Codebolt is built for that world.
+
+The major Features that allow Codebolt to handle such parallel agent Swarm are:
+- Custom Agents: In Codebolt developers have complete control over the Agent Logic and Agentic Loop. Unlike other applications where you do not have any control over the agent, and can just control the applicaiton using prompts, codebolt opens up the complete agent logic to the developer allowing the users to write any custom agents.
+- Remote Execution Agents: Codebolt allows 
+- Remote Environment Agents: 
+- Parallel Tasks:
+- Parallel Child Agents:
+- RunTime Task Generation:
+
+Also Codebolt Provides a lot of tools for users to manage Large Number of Agents running parallely using Swarm Management and Observervability tools. Some of the Tools that allow the users to manage such a large set of agents are:
+- Task Dashboard
+- Task Grouping
+- SubTask Hierarchy
+- Agent Monitoring
+- Remote Task Fanout
+- Task Status Monitoring
+
+These functionalities allow users to create different Strategies for multi-agent Swarm 
+- Local Agent Remote Executions
+- Parallel Remote Environemts
+- Hierarichal Remote Environments
+
+Codebolt is built for long duration agent executions spanning hours or days. The problem with long duration Agent Execution is that for single agent to lose the context or in case one agent makes a mistake in the starting, then that mistake becomes a big problem in long term. We have tools to help you handle these issues like:
+- Agent Grouping and Checking
+- Test and Critique agent
+- Custom Code Agents, to handle Isolated functionality
+
+Codebolt is also built for heavy parallelization. The tools that allow
+- Dynamic Task Creation by Agent
+- Hierarichal and Parallel Agent Execution
+
+Codebolt also Tries to provide optimize the Agent Functionality for Various roles by providing tools like:
+- Per Agent and Per task LLM Selection
+- Local LLM including fine-tuned llm selection.
+- Auto LLM Selection.