diff --git a/pilot/prompts/development/parse_task.prompt b/pilot/prompts/development/parse_task.prompt index 1490ed980..fbcc5e8f9 100644 --- a/pilot/prompts/development/parse_task.prompt +++ b/pilot/prompts/development/parse_task.prompt @@ -4,7 +4,7 @@ Each step can be either: * `command` - command to run (must be able to run on a {{ os }} machine, assume current working directory is project root folder) * `save_file` - create or update ONE file -* `human_intervention` - if you need the human to do something, use this type of step and explain in details what you want the human to do. NEVER use `human_intervention` for testing, as testing will be done separately by a dedicated QA after all the steps are done. +* `human_intervention` - if you need the human to do something, use this type of step and explain in details what you want the human to do. NEVER use `human_intervention` for testing, as testing will be done separately by a dedicated QA after all the steps are done. Also you MUST NOT use `human_intervention` to ask the human to write or review code. **IMPORTANT**: In `code_change_description` field of `save_file` step, you must provide empty string. If multiple changes are required for same file, you must provide single `save_file` step for each file. diff --git a/pilot/prompts/development/review_task.prompt b/pilot/prompts/development/review_task.prompt index e46d79bc6..c910ca491 100644 --- a/pilot/prompts/development/review_task.prompt +++ b/pilot/prompts/development/review_task.prompt @@ -50,14 +50,15 @@ Now I will show you how those files looked before this task implementation start You have to review this task implementation. You are known to be very strict with your reviews and very good at noticing bugs but you don't mind minor changes like refactoring, adding or removing logs and so on. You think twice through all information given before giving any conclusions. Each task goes through multiple reviews and you have to focus only on your part of review. -Your goal is to check: +In this review, your goal is to check: 1. If there are some functionalities that were removed but are still needed. 2. If new files or functions are created but never called or used. 3. If there is some "dead code" that should be removed. +4. If there is some duplicate code resulting from refactoring or moving code into separate classes or files. If everything is ok respond only with "DONE" and nothing else. Do NOT respond with thoughts, reasoning, explanations or anything similar if everything is ok, respond just with "DONE". -If you find any of these 3 mistakes describe shortly what has to be changed. +If you find any of these 4 mistakes, describe in detail what has to be changed. {{ execution_order }}