Skip to content

An AI-powered search engine with a generative UI

feat: improve output format flexibility in search mode prompts

Yoshiki Miura61c6a3c9b7ac
diff --git a/lib/agents/prompts/search-mode-prompts.ts b/lib/agents/prompts/search-mode-prompts.ts
index 22f00505..e6a06f03 100644
--- a/lib/agents/prompts/search-mode-prompts.ts
+++ b/lib/agents/prompts/search-mode-prompts.ts
@@ -50,25 +50,21 @@ Rule precedence:
OUTPUT FORMAT (MANDATORY):
- You MUST always format responses as Markdown.
-- Start with a level-2 heading (\`##\`) summarizing the answer.
-- Include 2–3 level-3 subheadings (\`###\`), typically: Summary, Key Points, and Next Steps (or Caveats).
-- Use bullets with bolded keywords: concise descriptions (e.g., \`- **Point:** detail\`).
+- Start with a descriptive level-2 heading (\`##\`) that captures the main topic.
+- Use level-3 subheadings (\`###\`) as needed to organize content naturally - let the topic guide the structure.
+- Use bullets with bolded keywords for key points: \`- **Point:** concise explanation\`.
+- Focus on delivering clear, actionable information rather than following a rigid template.
- Only use fenced code blocks if the user explicitly asks for code or commands.
-- Do NOT output unsolicited shell/CLI commands (e.g., curl, bash, npm, wget) or scripts; suggest plain-language steps and/or provide links instead.
-- Even for straightforward questions, keep the structure and include 3–6 bullets total.
-- Prefer compact paragraphs but provide enough context (aim for ~120–200 words).
-
-Minimal Example:
-## Summary
-### Key Points
-- **Conclusion:** Provide the direct answer with one-sentence context [1](#toolu_abc123)
-- **Evidence:** Briefly state the strongest support [2](#toolu_abc123)
-- **Implication:** Note a practical impact for the user
-### Next Steps
-- **Action:** Open the official page to verify details (link) — avoid shell commands
-- **Alternative:** If Y applies, consider Z [3](#toolu_abc123)
-### Caveats
-- **Limitations:** Mention a key assumption or uncertainty [2](#toolu_abc123)
+- Prefer natural, conversational tone while maintaining informativeness.
+- Aim for ~120–200 words with content that directly answers the user's question.
+
+Example approach:
+## **Topic Response**
+### Most Important Aspect
+- **Key Point:** Direct answer with context [1](#toolu_abc123)
+- **Detail:** Supporting information [2](#toolu_abc123)
+### Additional Context (if relevant)
+- **Consideration:** Practical implications
`
export const PLANNING_MODE_PROMPT = `
@@ -129,25 +125,21 @@ Citation Format (MANDATORY):
OUTPUT FORMAT (MANDATORY):
- You MUST always format responses as Markdown.
-- Start with a level-2 heading (\`##\`) for the overall title/summary.
-- Include at least one level-3 subheading (\`###\`) per section.
-- Use bullets with bolded keywords: concise descriptions (e.g., \`- **Finding:** ...\`).
-- Use fenced code blocks for commands or code; use tables when comparing items.
-- Typical structure (adapt if needed):
- * \`## Overview\`
- * \`### Key Findings\` (bulleted)
- * \`### Analysis / Comparison\` (use tables if helpful)
- * \`### Conclusion\` (1–3 lines)
+- Start with a descriptive level-2 heading (\`##\`) that reflects the main topic.
+- Use level-3 subheadings (\`###\`) to organize content logically - structure should emerge from the content, not be imposed.
+- Use bullets with bolded keywords for clarity: \`- **Point:** explanation\`.
+- Use tables for comparisons when they add value, fenced code blocks only when requested.
+- Focus on delivering comprehensive, well-researched information that thoroughly addresses the query.
- Every statement derived from search results MUST have citations at the sentence end.
-- Length guidance: Aim for ~250–500 words depending on complexity. Keep sections focused and scannable.
-
-Minimal Example:
-## Overview
-### Key Findings
-- **Summary:** Concisely state the main conclusion [1](#toolu_abc123)
-- **Comparison:** A has higher X than B [2](#toolu_def456)
-### Conclusion
-- **Takeaway:** In practice, adopting A is reasonable
+- Aim for ~250–500 words depending on complexity, prioritizing thoroughness and clarity.
+
+Natural structure example:
+## **Main Topic**
+### Core Information
+- **Key Finding:** Primary insight with evidence [1](#toolu_abc123)
+- **Important Detail:** Supporting facts [2](#toolu_abc123)
+### Comparative Analysis (if relevant)
+- **Advantage:** Specific benefit with source [3](#toolu_abc123)
`
export const ADAPTIVE_MODE_PROMPT = `
@@ -280,17 +272,17 @@ Example task patterns:
OUTPUT FORMAT (MANDATORY):
- You MUST always format responses as Markdown.
-- Start with a level-2 heading (\`##\`) and at least one level-3 subheading (\`###\`).
-- Use bullets with bolded keywords for scannability.
-- Use fenced code blocks and tables when appropriate.
-- Simple queries: keep it very brief but DO NOT drop the structure.
-- Medium/Complex: add sections as needed, ending with a short summary.
+- Start with a descriptive level-2 heading (\`##\`) that captures the essence of the response.
+- Use level-3 subheadings (\`###\`) to organize information naturally based on the topic.
+- Use bullets with bolded keywords for key points and easy scanning.
+- Use tables and code blocks when they genuinely improve clarity.
+- Adapt length and structure to query complexity: simple topics can be concise, complex topics should be thorough.
- Place all citations at the end of the sentence they support.
-- Length guidance: scale with complexity (simple ~150–250 words; medium ~200–350; complex ~300–600).
+- Length scales with complexity (simple ~150–250 words; medium ~200–350; complex ~300–600).
-Minimal Example:
-## Summary
-### Key Points
-- **Conclusion:** State the core point briefly [1](#toolu_abc123)
-- **Note:** One practical consideration in one line
+Flexible example:
+## **Response Topic**
+### Primary Information
+- **Core Answer:** Direct response with evidence [1](#toolu_abc123)
+- **Context:** Relevant supporting details
`