feat: Add Agents Setup and Documentation Sync (#109)

backend/AGENTS.md

@@ -0,0 +1,7 @@
+# Agent Workflow and Configuration Guidelines
+
+This document details the standardized configuration matrix, deployment patterns, and operational runtime conventions for autonomous and predictive tracking agents within PatchPilot.
+
+## Configuration Profiles
+* **Execution Strategy**: Define fallback pipelines for environment-isolated agent blocks.
+* **Skill Integration**: Standardize runtime tool payloads for local context processing.

backend/app/agents.py

@@ -0,0 +1,26 @@
+import os
+from typing import Dict, Any
+
+class PatchPilotAgentManager:
+    """Manages system configurations and operational parameters for AI tools/skills."""
+
+    def __init__(self):
+        self.enabled_skills: Dict[str, Any] = {}
+        # Default fallback context configs
+        self.default_context = {
+            "isolation_mode": True,
+            "max_token_limit": 4096
+        }
+
+    def register_agent_skill(self, skill_name: str, config_payload: Dict[str, Any]) -> None:
+        """Registers and validates a runtime agent competency skill block."""
+        if not skill_name:
+            raise ValueError("Skill name cannot be empty.")
+        self.enabled_skills[skill_name] = {**self.default_context, **config_payload}
+
+    def get_active_skills(self) -> Dict[str, Any]:
+        """Returns the fully mapped ecosystem tracking configuration."""
+        return self.enabled_skills
+
+# Global management tracking object initialized for app integration
+agent_manager = PatchPilotAgentManager()

backend/requirements.txt

@@ -8,4 +8,5 @@ aiosqlite
 joblib
 pandas
 scikit-learn
-starlette<1.0.0
+starlette<1.0.0
+langchain-core