Skip to content

Remove stale text initial maze prompt strategy#26

Closed
helenlu66 wants to merge 1 commit into
mainfrom
removed_unused_imports
Closed

Remove stale text initial maze prompt strategy#26
helenlu66 wants to merge 1 commit into
mainfrom
removed_unused_imports

Conversation

@helenlu66

Copy link
Copy Markdown
Member

Summary

  • remove the stale text_initial_maze prompting option
  • delete the unused TextInitialMazePromptStrategy
  • clean up now-unused runner imports and initial-maze prompt plumbing

Tests

  • python3 -m pytest tests/test_prompt_observation_text.py

@helenlu66

Copy link
Copy Markdown
Member Author

Closing per request; the stale prompt strategy removals are being undone.

@helenlu66 helenlu66 closed this Jul 2, 2026

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Removes the deprecated text_initial_maze prompting option and associated prompt strategy, simplifying prompt-strategy selection and runner prompt plumbing.

Changes:

  • Dropped text_initial_maze from ExperimentConfig.prompting and removed its strategy class.
  • Simplified ExperimentRunner prompt strategy mapping and system-prompt construction.
  • Reduced runner logging and transcript metadata (notably token-usage capture).

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File Description
interface/runner.py Removes text_initial_maze strategy wiring and simplifies system-prompt / transcript logging.
interface/prompt_strategies.py Deletes TextInitialMazePromptStrategy.
interface/config.py Removes text_initial_maze from the prompting Literal type.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread interface/runner.py
Comment on lines 127 to 131
def build_prompt_message(
self,
state,
last_feedback: str,
transcript: List[dict],
Comment thread interface/runner.py
Comment on lines 206 to 207
t_llm = time.perf_counter()
model_text = agent(agent_messages)
Comment thread interface/runner.py
Comment on lines +224 to +239
transcript.append(
{
"kind": "query",
"query_index": query_count,
"env_step_count": state.step_count,
"agent_messages": copy.deepcopy(agent_messages),
"assistant_reply": model_text,
"parsed_actions": list(action_queue),
"parse_ok": bool(action_queue),
"has_image": has_image,
"llm_latency_s": llm_s,
"chat_history_mode": chat_history,
"agent_message_count": len(agent_messages),
"actions_remaining_before_step": len(action_queue),
}
)
@helenlu66

Copy link
Copy Markdown
Member Author

Reopening with the requested approach: keep stale prompt strategy removal, but restore initial maze text in ExperimentRunner._build_message for text_only and image_text.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants