Parse the planner's sub-question list

Background

The Planner returns a numbered list of sub-questions as plain text. The agent parses it into a Python list: each line that starts with a digit and contains ". " becomes a sub-question (with the "N. " prefix removed).

for line in text.strip().splitlines():
    line = line.strip()
    if line and line[0].isdigit() and ". " in line:
        questions.append(line.split(". ", 1)[1].strip())

Problem statement

Implement parse_plan(text) returning the list of sub-question strings.

Input

• text — the planner's raw numbered-list output.

Output

Returns a list of strings: the text after "N. " for each numbered line, in order. Non-numbered lines are ignored.

Examples

Example 1

Input:  "1. What benchmark did AlphaCode use?\n2. What was its training data?"
Output: ["What benchmark did AlphaCode use?", "What was its training data?"]

Example 2 — preamble ignored

Input:  "Here is the plan:\n1. First\n2. Second"
Output: ["First", "Second"]

Constraints

• Keep only lines whose first character is a digit and that contain ". ".
• Strip the "N. " prefix and surrounding whitespace.
• Preserve order; skip blank/non-numbered lines.

Notes

• A clean list is the agent's checklist — every item becomes one executor, guaranteeing coverage.