Simplifying prompt newlines to use --- hrule and removed extra hrules

[jamesbraza]

--- is a basic Markdown hrule (link), which is also enforced by prettier in actual .md files.

This PR cleans up our prompting a bunch:

1. Moves our prompts to match this hrule convention, which has a few benefits:
   • Our prompt rendering better aligns with Markdown renderers
     • This may benefit how models perform on our prompts
   • Less prompt tokens
2. Inserts a placeholder (no text) if a Context has empty string text, to avoid model confusion.
3. Removes extra --- on table creation
4. Removes extra whitespace from qa_prompt

After this PR, here is a Context creation prompt involving a table:

Summarize the excerpt below to help answer a question.

Excerpt from Meeson2019 pages 25-31: Meeson, Richard, et al. "The Influence of Gap Size on the Development of Fracture Union with a Micro External Fixator." *Journal of the Mechanical Behavior of Biomedical Materials*, vol. 97, 2019, doi:10.1016/j.jmbbm.2019.07.015. Accessed 2025.

---

Placeholder

---

Markdown tables from Meeson2019 pages 25-31: Meeson, Richard, et al. "The Influence of Gap Size on the Development of Fracture Union with a Micro External Fixator." *Journal of the Mechanical Behavior of Biomedical Materials*, vol. 97, 2019, doi:10.1016/j.jmbbm.2019.07.015. Accessed 2025.. If the markdown is poorly formatted, defer to the images.

|% TISSUE|1.0mm Gap|1.5mm Gap|2.0mm Gap|
|---|---|---|---|
|**Bone**|45.6±33.0|39.1±23.9|23.2±26.6|
|**Cartilage**|36.7±22.1|43.1±24.6|37.2±17.9|
|**Fibrous**|14.7±30.6|15.3±37.4|36.1±40.8|
|**Vascular**<br>|3.0±1.9|2.4±2.0|3.5±3.4|

---

Question: What osteotomy gap (mm) has the bone volume per slice?

Do not directly answer the question, instead summarize to give evidence to help answer the question. Stay detailed; report specific numbers, equations, or direct quotes (marked with quotation marks). Reply "Not applicable" if the excerpt is irrelevant. At the end of your response, provide an integer score from 1-10 on a newline indicating relevance to question. Do not explain your score.

Relevant Information Summary (25 to 50 words):

[Copilot]

Pull Request Overview

This PR standardizes horizontal rule formatting in prompts by replacing ---- and ------------ with the standard Markdown --- hrule, reducing token usage and improving compatibility with Markdown renderers. Additionally, it adds (no text) placeholder when context text is empty to prevent model confusion.

Key changes:

• Standardized all horizontal rules to --- format
• Added placeholder text for empty contexts
• Updated test assertions to match new separator format

Reviewed Changes

Copilot reviewed 5 out of 12 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
tests/test_paperqa.py	Updated test assertion to expect --- separator instead of ----
tests/cassettes/test_partly_embedded_texts[True].yaml	Test cassette updated with new prompt format and API version changes
src/paperqa/settings.py	Changed context section separator from ---- to ---
src/paperqa/prompts.py	Standardized all horizontal rules to --- and reorganized prompt templates
src/paperqa/core.py	Added (no text) placeholder for empty text and removed extra separator in table formatting

---

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

[mskarlin]

lgtm