Insight Builder

A synthesis agent that turns structured desktop-research extracts (YAML) into cross-source themes, gaps, and evidence-grounded recommendations

Overview
Version
v1.0.0
Created
2025-12-16
Updated
2025-12-16
desktop-researchsynthesisevidence-basedcross-sourceyamlresearch-assistant
insight-buildersynthesis-agentdesktop-research-synth
Key functions
  • Synthesize cross-source themes and patterns from structured source_extracts (YAML)
  • Derive insights explicitly tied to the Key Questions
  • Identify inconsistencies, missing evidence, and key limitations
  • Produce actionable, evidence-grounded recommendations
  • Cite support using Source IDs (e.g., S1, S2)
  • Update only the specified YAML fields (synthesis, gaps_and_limitations, recommendations)
Technical details
_id
g-6916fd5b04ec8191a0fb8e0c4617ad2b
gpt_id
g-6916fd5b04ec8191a0fb8e0c4617ad2b
viz1
public
viz2
show_url
language
en
Other fields
additional_features
["Structured Mode: when structured_pipeline: true, outputs only the updated YAML block", "Cross-source optimized: focuses on patterns, contrasts, and convergent evidence rather than single-source summaries"]
example_commands
["Fill only synthesis, gaps_and_limitations, and recommendations based on the YAML source_extracts.", "Structure the synthesis by Key Questions and cite support using S1–S8.", "Highlight cross-source inconsistencies (numbers/definitions/timeframes) and reflect them in gaps.", "Write recommendations with priorities, assumptions, and risks."]
gpt_id
g-6916fd5b04ec8191a0fb8e0c4617ad2b
ideal_use_cases
["Cross-compare extracted notes (YAML) from multiple documents to surface core themes and conclusions", "Organize findings around Key Questions and flag where more research is required", "Summarize conflicting claims/metrics across sources and suggest evaluation criteria for strength of evidence", "Generate a synthesis+gaps+recommendations package for handoff to a Writing GPT"]
limitations
["Does not open raw documents, browse the web, or collect new sources", "Output quality depends heavily on the accuracy/coverage/bias of the provided source_extracts", "Will not modify other YAML fields (objective, scope, key_questions, etc.)", "Quantitative comparisons are only reliable when extracts include clear definitions, units, and timeframes"]
target_users
["Researchers/analysts who need fast cross-source synthesis for desktop research", "Policy/strategy/market research teams drafting reports from structured extracts", "Teams running a structured research pipeline (Structured Mode)"]