tony/hive
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

🎉 Complete CCLI Integration: Phase 4 (MCP Server Updates)

Browse Source 
IMPLEMENTATION COMPLETE: Successfully integrated Google Gemini CLI as
mixed agent type in Hive distributed AI platform.

## Phase 4 Achievements:
✅ Enhanced MCP tools with CLI agent support
✅ Added hive_register_cli_agent, hive_get_cli_agents tools
✅ Updated HiveClient interface for CLI agent management
✅ Mixed agent type coordination via MCP
✅ Comprehensive error handling and user feedback

## Key Features:
- CLI agent registration with health checks
- Mixed agent dashboard (🤖 Ollama + ⚡ CLI)
- Predefined agent quick setup (walnut-gemini, ironwood-gemini)
- SSH-based task execution with connection pooling
- Complete backward compatibility

## Technical Stack:
- MCP Tools: CLI agent management interface
- HiveClient: Enhanced API client with CLI support
- TypeScript: Full type safety for mixed agent operations
- Error Handling: Comprehensive CLI connectivity validation

## Production Ready:
✅ 16 MCP tools with CLI agent coverage
✅ Mixed agent type task coordination
✅ Health monitoring and statistics collection
✅ Robust SSH execution with timeout handling
✅ Integration tested and validated

Ready for hybrid AI orchestration: 5 Ollama + 2 CLI agents

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
anthonyrawlins
2025-07-10 12:11:27 +10:00
parent 85bf1341f3
commit 2915ee9aa7
9 changed files with 1045 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files

161
backend/app/core/hive_coordinator.py
View File

@@ -14,6 +14,7 @@ from enum import Enum
from sqlalchemy.orm import Session
from ..models.agent import Agent as ORMAgent
from ..core.database import SessionLocal
from ..cli_agents.cli_agent_manager import get_cli_agent_manager
class AgentType(Enum):
KERNEL_DEV = "kernel_dev"
@@ -21,6 +22,10 @@ class AgentType(Enum):
PROFILER = "profiler"
DOCS_WRITER = "docs_writer"
TESTER = "tester"
# CLI Agent Types
CLI_GEMINI = "cli_gemini"
GENERAL_AI = "general_ai"
REASONING = "reasoning"
class TaskStatus(Enum):
PENDING = "pending"
@@ -36,6 +41,8 @@ class Agent:
specialty: AgentType
max_concurrent: int = 2
current_tasks: int = 0
agent_type: str = "ollama" # "ollama" or "cli"
cli_config: Optional[Dict[str, Any]] = None
@dataclass
class Task:
@@ -56,6 +63,7 @@ class HiveCoordinator:
self.tasks: Dict[str, Task] = {}
self.task_queue: List[Task] = []
self.is_initialized = False
self.cli_agent_manager = None
# Agent prompts with compressed notation for efficient inter-agent communication
self.agent_prompts = {
@@ -87,7 +95,26 @@ FOCUS:[clear-accurate]→[explain+demonstrate+guide+solve]""",
SPEC:[coverage+benchmarks+edge-cases+automation]→[comprehensive+automated]
OUT:[tests+benchmarks+edge_cases+ci_config]→JSON[tests|benchmarks|edge_cases|ci_config]
FOCUS:[full-coverage]→[test+measure+handle+automate]"""
FOCUS:[full-coverage]→[test+measure+handle+automate]""",
# CLI Agent Prompts
AgentType.CLI_GEMINI: """[AI-assistant]→[general-purpose+reasoning]|[Gemini-2.5-Pro]
SPEC:[comprehensive-analysis+structured-responses+context-aware]→[accurate+helpful+efficient]
OUT:[detailed-response+reasoning+recommendations]→JSON[response|analysis|recommendations]
FOCUS:[intelligent-assistance]→[understand+analyze+explain+assist]""",
AgentType.GENERAL_AI: """[general-AI]→[multi-domain+adaptive]|[broad-knowledge]
SPEC:[flexible-reasoning+cross-domain+context-adaptation]→[accurate+comprehensive]
OUT:[detailed-analysis+insights+recommendations]→JSON[response|insights|recommendations]
FOCUS:[adaptive-intelligence]→[understand+reason+synthesize+recommend]""",
AgentType.REASONING: """[reasoning-expert]→[logical-analysis+problem-solving]|[advanced-reasoning]
SPEC:[complex-reasoning+step-by-step+logical-deduction]→[clear+systematic+thorough]
OUT:[detailed-reasoning+step-by-step+conclusions]→JSON[reasoning|steps|conclusions]
FOCUS:[logical-analysis]→[analyze+deduce+explain+conclude]"""
}
def add_agent(self, agent: Agent):
@@ -99,11 +126,25 @@ FOCUS:[full-coverage]→[test+measure+handle+automate]"""
model=agent.model,
specialty=agent.specialty.value,
max_concurrent=agent.max_concurrent,
current_tasks=agent.current_tasks
current_tasks=agent.current_tasks,
agent_type=agent.agent_type,
cli_config=agent.cli_config
)
db.add(db_agent)
db.commit()
db.refresh(db_agent)
# If it's a CLI agent, register with CLI agent manager
if agent.agent_type == "cli" and agent.cli_config:
try:
if not self.cli_agent_manager:
self.cli_agent_manager = get_cli_agent_manager()
self.cli_agent_manager.create_cli_agent(agent.id, agent.cli_config)
print(f"Registered CLI agent {agent.id} ({agent.specialty.value}) and initialized CLI backend")
except Exception as e:
print(f"Warning: Failed to initialize CLI backend for {agent.id}: {e}")
print(f"Registered agent {agent.id} ({agent.specialty.value}) at {agent.endpoint} and persisted to DB")
def create_task(self, task_type: AgentType, context: Dict, priority: int = 3) -> Task:
@@ -140,7 +181,9 @@ FOCUS:[full-coverage]→[test+measure+handle+automate]"""
model=db_agent.model,
specialty=AgentType(db_agent.specialty),
max_concurrent=db_agent.max_concurrent,
current_tasks=db_agent.current_tasks
current_tasks=db_agent.current_tasks,
agent_type=db_agent.agent_type or "ollama",
cli_config=db_agent.cli_config
)
return None
@@ -158,6 +201,46 @@ FOCUS:[full-coverage]→[test+measure+handle+automate]"""
task.status = TaskStatus.IN_PROGRESS
task.assigned_agent = agent.id
try:
# Route to appropriate executor based on agent type
if agent.agent_type == "cli":
result = await self._execute_cli_task(task, agent)
else:
result = await self._execute_ollama_task(task, agent)
task.result = result
task.status = TaskStatus.COMPLETED
task.completed_at = time.time()
print(f"Task {task.id} completed by {agent.id}")
return result
except Exception as e:
task.status = TaskStatus.FAILED
task.result = {"error": str(e)}
print(f"Task {task.id} failed: {e}")
return {"error": str(e)}
finally:
with SessionLocal() as db:
db_agent = db.query(ORMAgent).filter(ORMAgent.id == agent.id).first()
if db_agent:
db_agent.current_tasks -= 1
db.add(db_agent)
db.commit()
db.refresh(db_agent)
agent.current_tasks = db_agent.current_tasks # Update in-memory object
async def _execute_cli_task(self, task: Task, agent: Agent) -> Dict:
"""Execute task on CLI agent"""
if not self.cli_agent_manager:
self.cli_agent_manager = get_cli_agent_manager()
if not self.cli_agent_manager.is_initialized:
await self.cli_agent_manager.initialize()
return await self.cli_agent_manager.execute_cli_task(agent.id, task)
async def _execute_ollama_task(self, task: Task, agent: Agent) -> Dict:
"""Execute task on Ollama agent (original implementation)"""
prompt = self.agent_prompts[task.type]
# Construct compressed context using terse notation
@@ -179,48 +262,22 @@ Complete task → respond JSON format specified above."""
}
}
try:
# Use the session initialized in the coordinator
session = getattr(self, 'session', None)
if not session:
raise Exception("HTTP session not initialized")
async with session.post(
f"{agent.endpoint}/api/generate",
json=payload,
timeout=aiohttp.ClientTimeout(total=300) # 5 minute timeout for AI tasks
) as response:
if response.status == 200:
result = await response.json()
task.result = result
task.status = TaskStatus.COMPLETED
task.completed_at = time.time()
print(f"Task {task.id} completed by {agent.id}")
return result
else:
error_text = await response.text()
raise Exception(f"HTTP {response.status}: {error_text}")
# Use the session initialized in the coordinator
session = getattr(self, 'session', None)
if not session:
raise Exception("HTTP session not initialized")
except asyncio.TimeoutError:
task.status = TaskStatus.FAILED
task.result = {"error": "Task execution timeout"}
print(f"Task {task.id} timed out on {agent.id}")
return {"error": "Task execution timeout"}
except Exception as e:
task.status = TaskStatus.FAILED
task.result = {"error": str(e)}
print(f"Task {task.id} failed: {e}")
return {"error": str(e)}
finally:
with SessionLocal() as db:
db_agent = db.query(ORMAgent).filter(ORMAgent.id == agent.id).first()
if db_agent:
db_agent.current_tasks -= 1
db.add(db_agent)
db.commit()
db.refresh(db_agent)
agent.current_tasks = db_agent.current_tasks # Update in-memory object
async with session.post(
f"{agent.endpoint}/api/generate",
json=payload,
timeout=aiohttp.ClientTimeout(total=300) # 5 minute timeout for AI tasks
) as response:
if response.status == 200:
result = await response.json()
return result
else:
error_text = await response.text()
raise Exception(f"HTTP {response.status}: {error_text}")
async def process_queue(self):
"""Process the task queue with available agents"""
@@ -335,6 +392,16 @@ Complete task → respond JSON format specified above."""
)
)
# Initialize CLI agent manager
try:
self.cli_agent_manager = get_cli_agent_manager()
await self.cli_agent_manager.initialize()
print("✅ CLI Agent Manager initialized")
except Exception as e:
print(f"⚠️ CLI Agent Manager initialization failed: {e}")
print("Continuing without CLI agents...")
self.cli_agent_manager = None
# Initialize task processing
self.task_processor = None
@@ -350,6 +417,8 @@ Complete task → respond JSON format specified above."""
# Clean up any partial initialization
if hasattr(self, 'session') and self.session:
await self.session.close()
if hasattr(self, 'cli_agent_manager') and self.cli_agent_manager:
await self.cli_agent_manager.shutdown()
raise
async def shutdown(self):
@@ -365,6 +434,10 @@ Complete task → respond JSON format specified above."""
task.status = TaskStatus.FAILED
task.result = {"error": "Coordinator shutdown"}
# Shutdown CLI agent manager
if hasattr(self, 'cli_agent_manager') and self.cli_agent_manager:
await self.cli_agent_manager.shutdown()
# Close HTTP session
if hasattr(self, 'session') and self.session:
await self.session.close()