feat: add assistant bot with LLM integration and configuration

- Implemented the assistant bot with basic command handling and LLM routing.
- Created configuration file for the assistant bot with personality, behavior, and LLM settings.
- Added system prompt for the assistant bot to define its capabilities and limitations.
- Developed registration script for creating Matrix bot users via Synapse admin API.
- Introduced common development scripts for agent management (start, stop, list, logs).
- Scaffolded new agent creation script to streamline the addition of new agents.
- Implemented agent removal script to disable agents without deleting data.

agents/assistant/agent.go

@@ -0,0 +1,41 @@
+// Package assistant defines the pure rules for the assistant bot.
+// Since this bot is primarily LLM-driven, most rules just route to the LLM.
+package assistant
+
+import (
+	"github.com/enmanuel/agents/pkg/decision"
+)
+
+// Rules returns the decision rules for the assistant bot.
+func Rules() []decision.Rule {
+	return []decision.Rule{
+		// !help — explicit help command
+		{
+			Name:  "help",
+			Match: decision.MatchCommand("help"),
+			Actions: []decision.Action{{
+				Kind: decision.ActionKindReply,
+				Reply: &decision.ReplyAction{
+					Content: "Soy tu asistente. Escríbeme directamente lo que necesitas:\n" +
+						"- Preguntas, explicaciones, resúmenes\n" +
+						"- Ayuda con código\n" +
+						"- Redacción de textos\n\n" +
+						"Para tareas de infraestructura, habla con @devops-bot.",
+				},
+			}},
+		},
+
+		// Any DM or mention → LLM
+		// This is the catch-all: if the message is a DM or a mention, send to LLM.
+		{
+			Name: "llm-all",
+			Match: func(ctx decision.MessageContext) bool {
+				return ctx.IsDirectMsg || ctx.IsMention
+			},
+			Actions: []decision.Action{{
+				Kind: decision.ActionKindLLM,
+				LLM:  &decision.LLMAction{},
+			}},
+		},
+	}
+}