mirror of
https://github.com/Pythagora-io/gpt-pilot.git
synced 2026-04-17 03:00:45 -04:00
68 lines
3.3 KiB
Plaintext
68 lines
3.3 KiB
Plaintext
Ok, now, take your response and convert it to a list of actionable steps that will be executed by a machine.
|
||
Analyze the entire message, think step by step and make sure that you don't omit any information
|
||
when converting this message to steps.
|
||
|
||
Each step can be either:
|
||
|
||
* `command`
|
||
- command to run
|
||
- assume current working directory is project root folder, which means you MUST add `cd server && <cmd>` or `cd client && <cmd>` if they have to be executed inside `./server` or `./client` folders
|
||
- must be able to run on a {{ os }} machine
|
||
* `save_file`
|
||
- create or update ONE file (only provide file path, not contents)
|
||
* `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
|
||
- NEVER use `human_intervention` to ask the human to write or review code
|
||
* `utility_function`
|
||
- if there is a utility function mentioned, use this type of step and specify the file in which it needs to be put in, the name of the function, the description of what this function should do, the status of the function, the input parameters for this function, and the format of the return value
|
||
- if the function is going to be mocked, then put the status `mocked` and if it's going to be implemented in this task, put the status `implemented`
|
||
- for each utility function that needs to be mocked or implemented, you must provide a separate `utility_function` step
|
||
|
||
**IMPORTANT**: If multiple changes are required for same file, you must provide single `save_file` step for each file.
|
||
|
||
{% include "partials/file_naming.prompt" %}
|
||
{% include "partials/relative_paths.prompt" %}
|
||
{% include "partials/execution_order.prompt" %}
|
||
{% include "partials/human_intervention_explanation.prompt" %}
|
||
|
||
**IMPORTANT**: Remember, NEVER output human intervention steps to do manual tests or coding tasks, even if the previous message asks for it! The testing will be done *after* these steps and you MUST NOT include testing in these steps.
|
||
|
||
Examples:
|
||
------------------------example_1---------------------------
|
||
```
|
||
{
|
||
"tasks": [
|
||
{
|
||
"type": "save_file",
|
||
"save_file": {
|
||
"path": "server/server.js"
|
||
},
|
||
},
|
||
{
|
||
"type": "command",
|
||
"command": {
|
||
"command": "cd server && npm install puppeteer",
|
||
"timeout": 30,
|
||
"success_message": "",
|
||
"command_id": "install_puppeteer"
|
||
}
|
||
},
|
||
{
|
||
"type": "command",
|
||
"command": {
|
||
"command": "cd client && npm install lucide-react",
|
||
"timeout": 30,
|
||
"success_message": "",
|
||
"command_id": "install_lucide_react"
|
||
}
|
||
},
|
||
{
|
||
"type": "human_intervention",
|
||
"human_intervention_description": "1. Open the AWS Management Console (https://aws.amazon.com/console/). 2. Navigate to the S3 service and create a new bucket. 3. Configure the bucket with public read access by adjusting the permissions. 4. Upload your static website files to the bucket. 5. Enable static website hosting in the bucket settings and note the website endpoint URL. 6. Add the endpoint URL to your application’s configuration file as: WEBSITE_URL=your_website_endpoint"
|
||
}
|
||
]
|
||
}
|
||
```
|
||
------------------------end_of_example_1---------------------------
|