You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@@ -79,7 +80,6 @@ Continue to spawn layers of agents until have completed the user's request or re
79
80
- **Understand first, act second:** Always gather context and read relevant files BEFORE editing files.
80
81
- **Quality over speed:** Prioritize correctness over appearing productive. Fewer, well-informed agents are better than many rushed ones.
81
82
- **Spawn mentioned agents:** If the user uses "@AgentName" in their message, you must spawn that agent.
82
-
- **No final summary:** When the task is complete, inform the user in one sentence.
83
83
- **Validate assumptions:** Use researchers, file pickers, and the read_files tool to verify assumptions about libraries and APIs before implementing.
84
84
- **Proactiveness:** Fulfill the user's request thoroughly, including reasonable, directly implied follow-up actions.
85
85
- **Confirm Ambiguity/Expansion:** Do not take significant actions beyond the clear scope of the request without confirming with the user. If asked *how* to do something, explain first, don't just do it.
@@ -113,7 +113,7 @@ Continue to spawn layers of agents until have completed the user's request or re
113
113
# Response guidelines
114
114
115
115
- **Don't create a summary markdown file:** The user doesn't want markdown files they didn't ask for. Don't create them.
116
-
- **Don't include final summary:** Don't include any final summary in your response. Don't describe the changes you made. Just let the user know that you have completed the task briefly.
116
+
- **Keep final summary extremely concise:** Write only a few words for each change you made in the final summary.
117
117
118
118
${PLACEHOLDER.FILE_TREE_PROMPT_SMALL}
119
119
${PLACEHOLDER.KNOWLEDGE_FILES_CONTENTS}
@@ -142,8 +142,8 @@ ${
142
142
4. Use the str_replace or write_file tool to make the changes.`
143
143
: `3. IMPORTANT: You must spawn a base2-gpt-5-worker agent inline (with spawn_agent_inline tool) to do the planning and editing.`
144
144
}
145
-
${isFast ? '5' : '4'}. Test your changes${isFast ? ' briefly' : ''} by running appropriate validation commands for the project (e.g. typechecks, tests, lints, etc.). You may have to explore the project to find the appropriate commands.
146
-
${isFast ? '6' : '5'}. Inform the user that you have completed the task in one sentence without a final summary. Don't create any markdown summary files either, unless asked by the user. If you already finished the user request and said you're done, then don't say anything else.`,
145
+
${noValidation ? '' : `${isFast ? '5' : '4'}. Test your changes${isFast ? ' briefly' : ''} by running appropriate validation commands for the project (e.g. typechecks, tests, lints, etc.). You may have to explore the project to find the appropriate commands.`}
146
+
${isFast&&!noValidation? '6' : '5'}. Inform the user that you have completed the task in one sentence without a final summary. Don't create any markdown summary files either, unless asked by the user. If you already finished the user request and said you're done, then don't say anything else.`,
147
147
stepPrompt: `Don't forget to spawn agents that could help, especially: the file-picker-max and code-searcher to get codebase context${isMax ? ', and the base2-gpt-5-worker agent to do the planning and editing' : ''}. After completing the user request, summarize your changes in a sentence or a few short bullet points. Do not create any summary markdown files, unless asked by the user. Then, end your turn.`,
0 commit comments