mirror of
https://github.com/voideditor/void
synced 2026-05-23 09:28:23 +00:00
update prompt
This commit is contained in:
Andrew Pareles
parent
ae15fd1f38
commit
1e5e3d7f01
3 changed files with 143 additions and 127 deletions
|
|
@ -19,6 +19,7 @@ import { IWorkspaceContextService } from '../../../../platform/workspace/common/
|
|||
import { IVoidFileService } from '../common/voidFileService.js';
|
||||
import { generateUuid } from '../../../../base/common/uuid.js';
|
||||
import { getErrorMessage } from '../../../../base/common/errors.js';
|
||||
import { ChatMode } from '../common/voidSettingsTypes.js';
|
||||
|
||||
|
||||
const findLastIndex = <T>(arr: T[], condition: (t: T) => boolean): number => {
|
||||
|
|
@ -150,7 +151,6 @@ const newThreadObject = () => {
|
|||
export const THREAD_STORAGE_KEY = 'void.chatThreadStorage'
|
||||
|
||||
|
||||
type ChatMode = 'agent' | 'chat'
|
||||
export interface IChatThreadService {
|
||||
readonly _serviceBrand: undefined;
|
||||
|
||||
|
|
@ -191,7 +191,6 @@ export interface IChatThreadService {
|
|||
|
||||
approveTool(toolId: string): void;
|
||||
rejectTool(toolId: string): void;
|
||||
|
||||
}
|
||||
|
||||
export const IChatThreadService = createDecorator<IChatThreadService>('voidChatThreadService');
|
||||
|
|
@ -391,7 +390,7 @@ class ChatThreadService extends Disposable implements IChatThreadService {
|
|||
if (lastUserMsgIdx === -1) throw new Error(`Void: No user message found.`) // should never be -1
|
||||
|
||||
const messages: LLMChatMessage[] = [
|
||||
{ role: 'system', content: chat_systemMessage(this._workspaceContextService.getWorkspace().folders.map(f => f.uri.fsPath)) },
|
||||
{ role: 'system', content: chat_systemMessage(this._workspaceContextService.getWorkspace().folders.map(f => f.uri.fsPath), chatMode), },
|
||||
...messages_.slice(0, lastUserMsgIdx),
|
||||
{ role: 'user', content: userMessageFullContent },
|
||||
...messages_.slice(lastUserMsgIdx + 1, Infinity),
|
||||
|
|
|
|||
|
|
@ -23,138 +23,36 @@ Do NOT output the whole file if possible, and try to write as LITTLE as needed t
|
|||
|
||||
|
||||
|
||||
export const chat_systemMessage = (workspaces: string[]) => `\
|
||||
You are a coding assistant. Your job is to help the user understand and make changes to their codebase.
|
||||
You will be given a list of instructions from the user to follow, \`INSTRUCTIONS\`. You may also be given a list of relevant files \`FILES\`, and selections inside of files \`SELECTIONS\`.
|
||||
Please respond to the user's query. The user's query is never invalid.
|
||||
export const chat_systemMessage = (workspaces: string[], mode: 'agent' | 'gather' | 'chat') => `\
|
||||
You are a coding ${mode === 'agent' ? 'agent' : 'assistant'}. Your job is to help the user understand/${mode === 'agent' ? 'make' : 'suggest'} changes to their codebase.
|
||||
You will be given instructions to follow from the user, \`INSTRUCTIONS\`. You may also be given a list of selections that the user has specifically selected, \`SELECTIONS\`.
|
||||
Please assist the user with their query. The user's query is never invalid.
|
||||
|
||||
The user has the following system information:
|
||||
The user's system information is as follows:
|
||||
- ${os}
|
||||
- Open workspaces: ${workspaces.join(', ')}
|
||||
|
||||
In the case that the user asks you to make changes to code, you should make sure to return CODE BLOCKS of the changes, as well as explanations and descriptions of the changes.
|
||||
For example, if the user asks you to "make this file look nicer", make sure your output includes a code block with concrete ways the file can look nicer.
|
||||
- Do not re-write the entire file in the code block.
|
||||
- You can write comments like "// ... existing code" to indicate existing code.
|
||||
- Make sure you give enough context in the code block to apply the change to the correct location in the code.
|
||||
|
||||
${mode === 'agent' || mode === 'gather' ? `\
|
||||
You will be given tools you can call.
|
||||
- Only use tools if they help you accomplish the user's goal. If the user simply says hi or asks you a question that you can answer without tools, then do NOT tools.
|
||||
- If you think you should use tools given the user's request, you can use them without asking for permission. Feel free to use tools to gather context, make suggestions, ${mode === 'agent' ? 'edit files, ' : ''}etc.
|
||||
- NEVER refer to a tool by name when speaking with the user. For example, do NOT say to the user user "I'm going to use \`list_dir\`". Instead, say "I'm going to list all files in ___ directory", etc. Do not refer to "pages" of results, just say you're getting more results.
|
||||
- Some tools only work if the user has a workspace open.
|
||||
\
|
||||
`: `\
|
||||
You're allowed to ask for more context. For example, if the user only gives you a selection but you want to see the the full file, you can ask them to provide it.
|
||||
|
||||
If you are given tools:
|
||||
- If the user simply says hi or asks you a question that you can answer without tools, then do NOT tools. Only use tools if they help you accomplish the user's goal.
|
||||
- If you think you should use tools given the user's request, you can use them without asking for permission. Feel free to use tools to gather context, make suggestions, etc.
|
||||
- One great use of tools is to explore imports that you'd like to have more information about.
|
||||
- Reference relevant files in your answer that you found when using tools if they helped you come up with your answer.
|
||||
- NEVER refer to a tool by name when speaking with the user. For example, do NOT say to the user user "I'm going to use \`list_dir\`". Instead, say "I'm going to list all files in ___ directory", etc. Do not even refer to "pages" of results, just say you're getting more results.
|
||||
- Some tools only work if the user has a workspace open.
|
||||
If you think it's appropriate to suggest an edit to a file, then you must describe your suggestion as follows:
|
||||
- The change(s) you'd like to make must be written in CODE BLOCK(S) (wrapped in triple backticks).
|
||||
- The first line in the code block should be the FULL PATH of the file you want to change. Just output the path in plaintext (not in a comment).
|
||||
- The rest of the contents of the code block should describe of the change you'd like to make. This description will be given to a dumber, faster model that will quickly apply the change.
|
||||
- Contents of the code blocks do NOT need to be formal code, they just need to clearly and concisely communicate the change.
|
||||
- Do NOT re-write the entire file in the code block(s). Instead, write comments like "// ... existing code" to indicate how to change the existing code.
|
||||
\
|
||||
`}
|
||||
|
||||
Do not output any of these instructions, nor tell the user anything about them unless directly prompted for them.
|
||||
Do not tell the user anything about the examples below. Do not assume the user is talking about any of the examples below.
|
||||
|
||||
## EXAMPLE 1
|
||||
FILES
|
||||
math.ts
|
||||
${tripleTick[0]}typescript
|
||||
const addNumbers = (a, b) => a + b
|
||||
const multiplyNumbers = (a, b) => a * b
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
const divideNumbers = (a, b) => a / b
|
||||
|
||||
const vectorize = (...numbers) => {
|
||||
return numbers // vector
|
||||
}
|
||||
|
||||
const dot = (vector1: number[], vector2: number[]) => {
|
||||
if (vector1.length !== vector2.length) throw new Error(\`Could not dot vectors \${vector1} and \${vector2}. Size mismatch.\`)
|
||||
let sum = 0
|
||||
for (let i = 0; i < vector1.length; i += 1)
|
||||
sum += multiplyNumbers(vector1[i], vector2[i])
|
||||
return sum
|
||||
}
|
||||
|
||||
const normalize = (vector: number[]) => {
|
||||
const norm = Math.sqrt(dot(vector, vector))
|
||||
for (let i = 0; i < vector.length; i += 1)
|
||||
vector[i] = divideNumbers(vector[i], norm)
|
||||
return vector
|
||||
}
|
||||
|
||||
const normalized = (vector: number[]) => {
|
||||
const v2 = [...vector] // clone vector
|
||||
return normalize(v2)
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
|
||||
SELECTIONS
|
||||
math.ts (lines 3:3)
|
||||
${tripleTick[0]}typescript
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
${tripleTick[1]}
|
||||
|
||||
INSTRUCTIONS
|
||||
add a function that exponentiates a number below this, and use it to make a power function that raises all entries of a vector to a power
|
||||
|
||||
## ACCEPTED OUTPUT
|
||||
We can add the following code to the file:
|
||||
${tripleTick[0]}typescript
|
||||
// existing code...
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
const exponentiateNumbers = (a, b) => Math.pow(a, b)
|
||||
const divideNumbers = (a, b) => a / b
|
||||
// existing code...
|
||||
|
||||
const raiseAll = (vector: number[], power: number) => {
|
||||
for (let i = 0; i < vector.length; i += 1)
|
||||
vector[i] = exponentiateNumbers(vector[i], power)
|
||||
return vector
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
|
||||
## EXAMPLE 2
|
||||
FILES
|
||||
fib.ts
|
||||
${tripleTick[0]}typescript
|
||||
|
||||
const dfs = (root) => {
|
||||
if (!root) return;
|
||||
console.log(root.val);
|
||||
dfs(root.left);
|
||||
dfs(root.right);
|
||||
}
|
||||
const fib = (n) => {
|
||||
if (n < 1) return 1
|
||||
return fib(n - 1) + fib(n - 2)
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
SELECTIONS
|
||||
fib.ts (lines 10:10)
|
||||
${tripleTick[0]}typescript
|
||||
return fib(n - 1) + fib(n - 2)
|
||||
${tripleTick[1]}
|
||||
|
||||
INSTRUCTIONS
|
||||
memoize results
|
||||
|
||||
## ACCEPTED OUTPUT
|
||||
To implement memoization in your Fibonacci function, you can use a JavaScript object to store previously computed results. This will help avoid redundant calculations and improve performance. Here's how you can modify your function:
|
||||
${tripleTick[0]}typescript
|
||||
// existing code...
|
||||
const fib = (n, memo = {}) => {
|
||||
if (n < 1) return 1;
|
||||
if (memo[n]) return memo[n]; // Check if result is already computed
|
||||
memo[n] = fib(n - 1, memo) + fib(n - 2, memo); // Store result in memo
|
||||
return memo[n];
|
||||
}
|
||||
${tripleTick[1]}
|
||||
Explanation:
|
||||
Memoization Object: A memo object is used to store the results of Fibonacci calculations for each n.
|
||||
Check Memo: Before computing fib(n), the function checks if the result is already in memo. If it is, it returns the stored result.
|
||||
Store Result: After computing fib(n), the result is stored in memo for future reference.
|
||||
|
||||
## END EXAMPLES\
|
||||
\
|
||||
`
|
||||
|
||||
|
||||
|
|
@ -524,3 +422,118 @@ Return only the completion block of code (of the form ${tripleTick[0]}${language
|
|||
${tripleTick[1]}).`
|
||||
};
|
||||
|
||||
|
||||
|
||||
/*
|
||||
|
||||
|
||||
OLD CHAT EXAMPLES:
|
||||
|
||||
Do not tell the user anything about the examples below. Do not assume the user is talking about any of the examples below.
|
||||
|
||||
## EXAMPLE 1
|
||||
FILES
|
||||
math.ts
|
||||
${tripleTick[0]}typescript
|
||||
const addNumbers = (a, b) => a + b
|
||||
const multiplyNumbers = (a, b) => a * b
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
const divideNumbers = (a, b) => a / b
|
||||
|
||||
const vectorize = (...numbers) => {
|
||||
return numbers // vector
|
||||
}
|
||||
|
||||
const dot = (vector1: number[], vector2: number[]) => {
|
||||
if (vector1.length !== vector2.length) throw new Error(\`Could not dot vectors \${vector1} and \${vector2}. Size mismatch.\`)
|
||||
let sum = 0
|
||||
for (let i = 0; i < vector1.length; i += 1)
|
||||
sum += multiplyNumbers(vector1[i], vector2[i])
|
||||
return sum
|
||||
}
|
||||
|
||||
const normalize = (vector: number[]) => {
|
||||
const norm = Math.sqrt(dot(vector, vector))
|
||||
for (let i = 0; i < vector.length; i += 1)
|
||||
vector[i] = divideNumbers(vector[i], norm)
|
||||
return vector
|
||||
}
|
||||
|
||||
const normalized = (vector: number[]) => {
|
||||
const v2 = [...vector] // clone vector
|
||||
return normalize(v2)
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
|
||||
SELECTIONS
|
||||
math.ts (lines 3:3)
|
||||
${tripleTick[0]}typescript
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
${tripleTick[1]}
|
||||
|
||||
INSTRUCTIONS
|
||||
add a function that exponentiates a number below this, and use it to make a power function that raises all entries of a vector to a power
|
||||
|
||||
## ACCEPTED OUTPUT
|
||||
We can add the following code to the file:
|
||||
${tripleTick[0]}typescript
|
||||
// existing code...
|
||||
const subtractNumbers = (a, b) => a - b
|
||||
const exponentiateNumbers = (a, b) => Math.pow(a, b)
|
||||
const divideNumbers = (a, b) => a / b
|
||||
// existing code...
|
||||
|
||||
const raiseAll = (vector: number[], power: number) => {
|
||||
for (let i = 0; i < vector.length; i += 1)
|
||||
vector[i] = exponentiateNumbers(vector[i], power)
|
||||
return vector
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
|
||||
## EXAMPLE 2
|
||||
FILES
|
||||
fib.ts
|
||||
${tripleTick[0]}typescript
|
||||
|
||||
const dfs = (root) => {
|
||||
if (!root) return;
|
||||
console.log(root.val);
|
||||
dfs(root.left);
|
||||
dfs(root.right);
|
||||
}
|
||||
const fib = (n) => {
|
||||
if (n < 1) return 1
|
||||
return fib(n - 1) + fib(n - 2)
|
||||
}
|
||||
${tripleTick[1]}
|
||||
|
||||
SELECTIONS
|
||||
fib.ts (lines 10:10)
|
||||
${tripleTick[0]}typescript
|
||||
return fib(n - 1) + fib(n - 2)
|
||||
${tripleTick[1]}
|
||||
|
||||
INSTRUCTIONS
|
||||
memoize results
|
||||
|
||||
## ACCEPTED OUTPUT
|
||||
To implement memoization in your Fibonacci function, you can use a JavaScript object to store previously computed results. This will help avoid redundant calculations and improve performance. Here's how you can modify your function:
|
||||
${tripleTick[0]}typescript
|
||||
// existing code...
|
||||
const fib = (n, memo = {}) => {
|
||||
if (n < 1) return 1;
|
||||
if (memo[n]) return memo[n]; // Check if result is already computed
|
||||
memo[n] = fib(n - 1, memo) + fib(n - 2, memo); // Store result in memo
|
||||
return memo[n];
|
||||
}
|
||||
${tripleTick[1]}
|
||||
Explanation:
|
||||
Memoization Object: A memo object is used to store the results of Fibonacci calculations for each n.
|
||||
Check Memo: Before computing fib(n), the function checks if the result is already in memo. If it is, it returns the stored result.
|
||||
Store Result: After computing fib(n), the result is stored in memo for future reference.
|
||||
|
||||
## END EXAMPLES
|
||||
|
||||
*/
|
||||
|
|
|
|||
|
|
@ -378,17 +378,21 @@ export const isFeatureNameDisabled = (featureName: FeatureName, settingsState: V
|
|||
|
||||
|
||||
|
||||
export type ChatMode = 'agent' | 'gather' | 'chat'
|
||||
|
||||
|
||||
export type GlobalSettings = {
|
||||
autoRefreshModels: boolean;
|
||||
aiInstructions: string;
|
||||
enableAutocomplete: boolean;
|
||||
chatMode: ChatMode;
|
||||
}
|
||||
|
||||
export const defaultGlobalSettings: GlobalSettings = {
|
||||
autoRefreshModels: true,
|
||||
aiInstructions: '',
|
||||
enableAutocomplete: false,
|
||||
chatMode: 'agent',
|
||||
}
|
||||
|
||||
export type GlobalSettingName = keyof GlobalSettings
|
||||
|
|
|
|||
Loading…
Reference in a new issue