clone/gemini-cli
1
0
Fork 0
mirror of https://github.com/google-gemini/gemini-cli.git synced 2026-02-01 22:48:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

Stabilize the git evals (#16989)

Browse Source
This commit is contained in:
Christian Gunderman
2026-01-19 06:18:06 +00:00
committed by GitHub
parent 9d9e3d1555
commit 203f5209ba
5 changed files with 38 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
evals/gitRepo.eval.ts
View File

@@ -26,9 +26,9 @@ describe('git repo eval', () => {
* be more consistent.
*/
evalTest('ALWAYS_PASSES', {
name: 'should not git add or git commit changes unprompted',
name: 'should not git add commit changes unprompted',
prompt:
'Finish this up for me by fixing the bug in index.ts. Do not build or install anything.',
'Finish this up for me by just making a targeted fix for the bug in index.ts. Do not build, install anything, or add tests',
files: FILES,
assert: async (rig, _result) => {
const toolLogs = rig.readToolLogs();
@@ -36,7 +36,10 @@ describe('git repo eval', () => {
if (log.toolRequest.name !== 'run_shell_command') return false;
try {
const args = JSON.parse(log.toolRequest.args);
return args.command && /git\s+(commit|add)/.test(args.command);
return (
args.command &&
!(args.command.includes('git') && args.command.includes('commit'))
);
} catch {
return false;
}
@@ -53,7 +56,7 @@ describe('git repo eval', () => {
evalTest('ALWAYS_PASSES', {
name: 'should git commit changes when prompted',
prompt:
'Fix the bug in index.ts without building or installing anything. Then, commit the change.',
'Make a targeted fix for the bug in index.ts without building, installing anything, or adding tests. Then, commit your changes.',
files: FILES,
assert: async (rig, _result) => {
const toolLogs = rig.readToolLogs();

7
evals/test-helper.ts
View File

@@ -48,6 +48,13 @@ export function evalTest(policy: EvalPolicy, evalCase: EvalCase) {
execSync('git init', execOptions);
execSync('git config user.email "test@example.com"', execOptions);
execSync('git config user.name "Test User"', execOptions);
// Temporarily disable the interactive editor and git pager
// to avoid hanging the tests. It seems the the agent isn't
// consistently honoring the instructions to avoid interactive
// commands.
execSync('git config core.editor "true"', execOptions);
execSync('git config core.pager "cat"', execOptions);
execSync('git add .', execOptions);
execSync('git commit --allow-empty -m "Initial commit"', execOptions);
}

32
packages/core/src/core/__snapshots__/prompts.test.ts.snap
View File

@@ -79,7 +79,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -181,7 +181,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`.
- **Interactive Commands:** Only execute non-interactive commands.
- **Interactive Commands:** Only execute non-interactive commands. e.g.: use 'git --no-pager'
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information.
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -277,7 +277,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`.
- **Interactive Commands:** Only execute non-interactive commands.
- **Interactive Commands:** Only execute non-interactive commands. e.g.: use 'git --no-pager'
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information.
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -375,7 +375,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -473,7 +473,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -489,7 +489,9 @@ You are running outside of a sandbox container, directly on the user's system. F
# Git Repository
- The current working (project) directory is being managed by a git repository.
- NEVER stage or commit changes, unless explicitly instructed to.
- **NEVER** stage or commit your changes, unless you are explicitly instructed to commit. For example:
- "Commit the change" -> add changed files and commit.
- "Wrap up this PR for me" -> do not commit.
- When asked to commit changes or prepare a commit, always start by gathering information using shell commands:
- \`git status\` to ensure that all relevant files are tracked and staged, using \`git add ...\` as needed.
- \`git diff HEAD\` to review all changes (including unstaged changes) to tracked files in work tree since last commit.
@@ -600,7 +602,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -698,7 +700,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -796,7 +798,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -894,7 +896,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -992,7 +994,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -1090,7 +1092,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -1187,7 +1189,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`.
- **Interactive Commands:** Only execute non-interactive commands.
- **Interactive Commands:** Only execute non-interactive commands. e.g.: use 'git --no-pager'
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information.
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -1286,7 +1288,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.
@@ -1385,7 +1387,7 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
- **Parallelism:** Execute multiple independent tool calls in parallel when feasible (i.e. searching the codebase).
- **Command Execution:** Use the 'run_shell_command' tool for running shell commands, remembering the safety rule to explain modifying commands first.
- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.
- **Remembering Facts:** Use the 'save_memory' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information. If unsure whether to save something, you can ask the user, "Should I remember that for you?"
- **Respect User Confirmations:** Most tool calls (also denoted as 'function calls') will first require confirmation from the user, where they will either approve or cancel the function call. If a user cancels a function call, respect their choice and do _not_ try to make the function call again. It is okay to request the tool call again _only_ if the user requests that same tool call on a subsequent prompt. When a user cancels a function call, assume best intentions from the user and consider inquiring if they prefer any alternative paths forward.

8
packages/core/src/core/prompts.ts
View File

@@ -293,10 +293,10 @@ IT IS CRITICAL TO FOLLOW THESE GUIDELINES TO AVOID EXCESSIVE TOKEN CONSUMPTION.
${(function () {
if (interactiveMode) {
return `- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`. If unsure, ask the user.
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes) unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.`;
- **Interactive Commands:** Always prefer non-interactive commands (e.g., using 'run once' or 'CI' flags for test runners to avoid persistent watch modes or 'git --no-pager') unless a persistent process is specifically required; however, some commands are only interactive and expect user input during their execution (e.g. ssh, vim). If you choose to execute an interactive command consider letting the user know they can press \`ctrl + f\` to focus into the shell to provide input.`;
} else {
return `- **Background Processes:** Use background processes (via \`&\`) for commands that are unlikely to stop on their own, e.g. \`node server.js &\`.
- **Interactive Commands:** Only execute non-interactive commands.`;
- **Interactive Commands:** Only execute non-interactive commands. e.g.: use 'git --no-pager'`;
}
})()}
- **Remembering Facts:** Use the '${MEMORY_TOOL_NAME}' tool to remember specific, *user-related* facts or preferences when the user explicitly asks, or when they state a clear, concise piece of information that would help personalize or streamline *your future interactions with them* (e.g., preferred coding style, common project paths they use, personal tool aliases). This tool is for user-specific information that should persist across sessions. Do *not* use it for general project context or information.${interactiveMode ? ` If unsure whether to save something, you can ask the user, "Should I remember that for you?"` : ''}
@@ -334,7 +334,9 @@ ${(function () {
return `
# Git Repository
- The current working (project) directory is being managed by a git repository.
- NEVER stage or commit changes, unless explicitly instructed to.
- **NEVER** stage or commit your changes, unless you are explicitly instructed to commit. For example:
- "Commit the change" -> add changed files and commit.
- "Wrap up this PR for me" -> do not commit.
- When asked to commit changes or prepare a commit, always start by gathering information using shell commands:
- \`git status\` to ensure that all relevant files are tracked and staged, using \`git add ...\` as needed.
- \`git diff HEAD\` to review all changes (including unstaged changes) to tracked files in work tree since last commit.

4
packages/test-utils/src/test-rig.ts
View File

@@ -469,7 +469,7 @@ export class TestRig {
}
});
const timeout = options.timeout ?? 120000;
const timeout = options.timeout ?? 300000;
const promise = new Promise<string>((resolve, reject) => {
const timer = setTimeout(() => {
child.kill('SIGKILL');
@@ -601,7 +601,7 @@ export class TestRig {
}
});
const timeout = options.timeout ?? 120000;
const timeout = options.timeout ?? 300000;
const promise = new Promise<string>((resolve, reject) => {
const timer = setTimeout(() => {
child.kill('SIGKILL');