synthesize-analysis

name	synthesize-analysis
description	Consolidate all available analysis for a single idea (lens outputs, council deliberation, comparisons) into one timestamped unified assessment with a clear verdict and recommended next step. Use after the user has run multiple lenses and wants a single document that pulls it all together.

Synthesize Analysis

Produces a unified, timestamped assessment that brings together every per-lens output for one idea.

Steps

1. Read the idea workspace <workspace_path>/ideas/<slug>/:
   • idea.md and any versions/*.md
   • All analysis/*.md
   • All council/*.md or council-karpathy/*.md if present
2. Build the synthesis as a single markdown document with these sections:
   • Idea summary — 2–3 sentence reduction of the captured idea
   • What we know — 1-paragraph distillation of each lens, naming each
   • Where the lenses agree — points of convergence across analyses
   • Where the lenses disagree — genuine tensions; do not resolve them artificially
   • The council's verdict — if council outputs exist, summarise the chairman synthesis and call out dissent
   • Top assumptions to validate next — pull from assumptions-and-objections.md if present
   • Recommended next step — one concrete action: a validation experiment, a build phase to start, or a kill decision
   • Verdict — one of: Strong Pursue / Pursue with Changes / Hold for Validation / Pass / Pivot Required
3. Tag the synthesis with timestamp so multiple runs accumulate (don't overwrite earlier syntheses).

Output

<workspace_path>/ideas/<slug>/synthesis/YYYY-MM-DD-HHMM.md with frontmatter:

---
type: idea-synthesis
generated: YYYY-MM-DD HH:MM
verdict: <one of the five>
inputs: [<list of source files folded in>]
---

Notes

• Always write to a new timestamped file. The history of syntheses is itself useful — it shows how thinking evolved.
• If only one or two lenses have been run, synthesise honestly with that limited base — don't pad.
• The recommended next step must be concrete enough to act on this week, not a vague "do more research".