agentica-claude-proxy

Use this skill when developing or debugging the Agentica-Claude proxy integration. Claude CLI in -p mode restricts file operations. You MUST add: Without this, agents will report "permission denied" for Write/Edit operations. Agentica expects SSE streaming, not plain JSON: Agents MUST return results as Python code blocks with a return statement: Agentica's REPL parser extracts code between ```python and ```. Agents will hallucinate success without actually using tools unless you explicitly warn them: Both Claude Code and Agentica have sandboxes: Solution: Use project-relative paths like workspace/ instead of /tmp/ Note: Logs only show final conversational response, not tool invocations. - Setting up Agentica agents to use Claude Code tools - Debugging agent hallucination issues - Fixing permission errors in file operations - Understanding the REPL response format - /tmp/ paths are blocked by Claude Code - Files outside project directory blocked by Agentica - Proxy implementation: scripts/agentica/claude_proxy.py - REPL_BASELINE prompt: scripts/agentica/claude_proxy.py:49-155 - Comprehensive test: workspace/test_swarm_all_tools.py - DependencySwarm: scripts/agentica/dependency_swarm.py