Case studies

Published-grade writeups showing what Sync does in concrete terms. Each case study takes a real Hermes interaction with a Sync-augmented agent and contrasts it against the same agent without Sync — same person, same question, same conversation context, one variable changed. The delta is the artifact.

Format and rationale live in docs/briefs/case-study-writeups-brief.md.

Contents

• sync-effect-self-comparison.md — Draft, 2026-05-25. Multi-comparison validation batch (6 prompts spanning strategic, technical, timing, open-ended, and bad-framed questions). All 6 showed structural value from Sync; Entry 5 (open-ended leverage) flipped the recommendation entirely. Spin-off single-comparison publishable version sketched at the bottom.

Adding a new case study

Follow the brief's format. Default structure: subject and context → setup → without-Sync response (verbatim) → with-Sync response (verbatim) → the delta → what this generalizes to. Keep both responses verbatim — no editing for length.

For multi-comparison batches (like the first one above), adapt the format but preserve the verbatim discipline. The point of the artifact is the side-by-side; abstracting it away defeats the purpose.