Robotframework: Self Healing

AI
1

Hi Team,

Need to check if we have any solution around self healing:

  1. During execution if locator fails, script should rectify and proceed: share report for the changes done.
  2. If any changes identified in flow, it should handle and share report.

If have tried selfhealing library with local ollama, but it takes long time for each object approx 5mins to execute.

2

Hi @SeemaT ,
I recommend to try out GitHub - MarketSquare/robotframework-selfhealing-agents: A robotframework library that repairs failing Robot Framework tests automatically via AI · GitHub

But: Use the current prerelease (0.2.0dev1) if you want to use local LLMS via Ollama.
So install via
pip install –pre robotframework-selfhealing-agents

I will do an official release in the next days.

1 Like
3

Thanks a lot.

But I want to setup local ollama, for which we don’t have any API keys.

How can we configure it?

Also can you help to understand in detail this architecture.

4

I assume you have already installed Ollama and some models locally.
Install the prerelease version 0.2.0dev1 from robotframework-selfhealing-agents.
Add a .env file to your project root.

LOCATOR_AGENT_PROVIDER=ollama
LOCATOR_AGENT_MODEL=qwen3:latest
ORCHESTRATOR_AGENT_PROVIDER=“ollama”
ORCHESTRATOR_AGENT_MODEL=“qwen3:latest”
BASE_URL=http://localhost:11434/v1
STRICT_TOOL_DEFINITIONS=false

ENABLE_SELF_HEALING=True
USE_LLM_FOR_LOCATOR_GENERATION=True
MAX_RETRIES=3
REQUEST_LIMIT=5
TOTAL_TOKENS_LIMIT=16000
IS_RERUN_ACTIVATED=False

You can find more infoes here in this branch:

1 Like