Can't add action

[maxenergy]

Dear Stoerr,
I fllow your steps , but in adding action, test it can not work.
my yaml file

http://localhost:7364/codeveloperengine.yaml

openapi: 3.0.1
info:
title: Co-Developer GPT Engine
version: THEVERSION
**servers:

• url: https://maxenergyluo.loca.lt/**
  paths:
  /executeAction:
  post:
  operationId: executeAction
  x-openai-isConsequential: false
  summary: Execute an action with given content as standard input. Only on explicit user request.
  parameters:
  - name: actionName
  in: query
  required: true
  schema:
  type: string
  requestBody:
  required: true
  content:
  application/json:
  schema:
  type: object
  properties:
  actionInput:
  type: string
  responses:
  '200':
  description: Action executed successfully, output returned
  content:
  text/plain:
  schema:
  type: string
  /fetchUrlTextContent:
  get:
  operationId: fetchUrlTextContent
  x-openai-isConsequential: false
  summary: Fetch text content from a given URL.
  parameters:
  - name: url
  in: query
  required: true
  schema:
  type: string
  - name: raw
  description: return raw html or pdf content without converting to markdown
  in: query
  required: false
  schema:
  type: boolean
  responses:
  '200':
  description: fetch successful; content returned
  content:
  text/plain:
  schema:
  type: string
  /grepFiles:
  get:
  operationId: grepAction
  x-openai-isConsequential: false
  summary: Search for lines in text files matching the given regex.
  parameters:
  - name: path
  in: query
  description: relative path to the directory to search in or the file to search. root directory = '.'
  required: true
  schema:
  type: string
  - name: fileRegex
  in: query
  description: optional regex to filter file names
  required: false
  schema:
  type: string
  - name: grepRegex
  in: query
  description: regex to filter lines in the files
  required: true
  schema:
  type: string
  - name: contextLines
  in: query
  description: number of context lines to include with each match (not yet used)
  required: false
  schema:
  type: integer
  responses:
  '200':
  description: Lines matching the regex
  content:
  text/plain:
  schema:
  type: string
  /listFiles:
  get:
  operationId: listFiles
  x-openai-isConsequential: false
  summary: Recursively lists files in a directory. Optionally filters by filename and content.
  parameters:
  - name: path
  in: query
  description: relative path to directory to list. root directory = '.'
  required: true
  schema:
  type: string
  - name: filePathRegex
  in: query
  description: regex to filter file paths - use for search by file name
  required: false
  schema:
  type: string
  - name: grepRegex
  in: query
  description: an optional regex that lists only files that contain a line matching this pattern
  required: false
  schema:
  type: string
  - name: listDirectories
  in: query
  description: if true, lists directories instead of files
  required: false
  schema:
  type: boolean
  responses:
  '200':
  description: List of relative paths of the files
  content:
  text/plain:
  schema:
  type: string
  /readFile:
  get:
  operationId: readFile
  x-openai-isConsequential: false
  summary: Read a files content.
  parameters:
  - name: path
  in: query
  description: relative path to file
  required: true
  schema:
  type: string
  - name: maxLines
  in: query
  description: maximum number of lines to read from the file
  required: false
  schema:
  type: integer
  - name: startLine
  in: query
  description: line number to start reading from
  required: false
  schema:
  type: integer
  responses:
  '200':
  description: Content of the file
  content:
  text/plain:
  schema:
  type: string
  /replaceInFile:
  post:
  operationId: replaceInFile
  x-openai-isConsequential: false
  summary: Replaces the single occurrence of one or more literal strings in a file. The whole file content is matched, not line by line.
  parameters:
  - name: path
  in: query
  description: relative path to file
  required: true
  schema:
  type: string
  requestBody:
  required: true
  content:
  application/json:
  schema:
  type: object
  properties:
  replacements:
  type: array
  items:
  type: object
  properties:
  search:
  type: string
  description: The literal string to be replaced - can contain many lines, but please take care to find a small number of lines to replace. Everything that is replaced must be here. Prefer to match the whole line / several whole lines.
  replace:
  type: string
  description: Literal replacement, can contain several lines. Please observe the correct indentation.
  responses:
  '200':
  description: File updated successfully
  /writeFile:
  post:
  operationId: writeFile
  x-openai-isConsequential: false
  summary: Overwrite a small file with the complete content given in one step. You cannot append to a file or write parts or write parts - use replaceInFile for inserting parts.
  parameters:
  - name: path
  in: query
  description: relative path to file
  required: true
  schema:
  type: string
  requestBody:
  required: true
  content:
  application/json:
  schema:
  type: object
  properties:
  content:
  type: string
  responses:
  '200':
  description: File overwritten

[stoerr]

Hi!

Thanks for trying and reporting the problem!

I thought I had that working with loca.lt , but I cannot get it to work right now, too. :-( I can import the codeveloperengine.yaml, as you did, but the requests for e.g. listing files never arrive at my machine. My guess is that that mechanism that asks you for a password the first time you're calling the tunnel from the browser does also hit ChatGPT when it's trying to access your tunnel , in your case https://maxenergyluo.loca.lt/ , and obviously ChatGPT can't enter the password there. I currently don't know a workaround for that. I guess I'll have to remove that recommendation from the documentation for now. Perhaps something changed in the implementation there.

Could you please try serveo instead, as described on https://codevelopergptengine.stoerr.net/https.html ? That's what I'm using and that always worked. Please give me a thumbs up if that works for you.

Best regards,
Hans-Peter

[stoerr]

Well, I updated the documentation that localtunnel currently doesn't work. I can't do anything about that except reporting a bug with OpenAI - it shouldn't fail silently without any kind of error message. Sorry about that - I can only recommend using serveo for now.

[maxenergy]

Tks for reply ,using serveo can works fine now. and my yaml has now more actions just /grepFiles: /grepFiles: /readFile:
like this:

https://maxenergyluo.serveo.net/codeveloperengine.yaml

openapi: 3.0.1
info:
title: Co-Developer GPT Engine
version: THEVERSION
servers:

• url: https://maxenergyluo.serveo.net
  paths:
  /grepFiles:
  get:
  operationId: grepAction
  x-openai-isConsequential: false
  summary: Search for lines in text files matching the given regex.
  parameters:
  - name: path
  in: query
  description: relative path to the directory to search in or the file to search. root directory = '.'
  required: true
  schema:
  type: string
  - name: fileRegex
  in: query
  description: optional regex to filter file names
  required: false
  schema:
  type: string
  - name: grepRegex
  in: query
  description: regex to filter lines in the files
  required: true
  schema:
  type: string
  - name: contextLines
  in: query
  description: number of context lines to include with each match (not yet used)
  required: false
  schema:
  type: integer
  responses:
  '200':
  description: Lines matching the regex
  content:
  text/plain:
  schema:
  type: string
  /listFiles:
  get:
  operationId: listFiles
  x-openai-isConsequential: false
  summary: Recursively lists files in a directory. Optionally filters by filename and content.
  parameters:
  - name: path
  in: query
  description: relative path to directory to list. root directory = '.'
  required: true
  schema:
  type: string
  - name: filePathRegex
  in: query
  description: regex to filter file paths - use for search by file name
  required: false
  schema:
  type: string
  - name: grepRegex
  in: query
  description: an optional regex that lists only files that contain a line matching this pattern
  required: false
  schema:
  type: string
  - name: listDirectories
  in: query
  description: if true, lists directories instead of files
  required: false
  schema:
  type: boolean
  responses:
  '200':
  description: List of relative paths of the files
  content:
  text/plain:
  schema:
  type: string
  /readFile:
  get:
  operationId: readFile
  x-openai-isConsequential: false
  summary: Read a files content.
  parameters:
  - name: path
  in: query
  description: relative path to file
  required: true
  schema:
  type: string
  - name: maxLines
  in: query
  description: maximum number of lines to read from the file
  required: false
  schema:
  type: integer
  - name: startLine
  in: query
  description: line number to start reading from
  required: false
  schema:
  type: integer
  responses:
  '200':
  description: Content of the file
  content:
  text/plain:
  schema:
  type: string

[stoerr]

Hi! Just saw the ticket again - how are your experiences with the whole thing?