lamina-research-synthesis

Solid

Simulation synthesis — merge parallel actor-walk and walkthrough results into findings[]. Not affinity mapping human interviews.

Code & Development 65 stars 2 forks Updated today Apache-2.0

Install

View on GitHub

Quality Score: 82/100

Stars 20%
61
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
36
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# Simulation Synthesis (agent-native) After parallel verify subagents return, **merge evidence** into actionable `findings[]`. ## Framework 1. Collect persona-walk JSON, walkthrough captures, and accessibility evidence 2. Cluster by: invariant violation, permission gap, UX blocker, a11y issue 3. Deduplicate same root cause across actors 4. Rank: blocking build trust vs polish 5. Link each finding to `scenario_id`, `screen_id`, or invariant id ## Anti-patterns - **Quote theater** — fabricated user quotes from simulated sessions - **Insight without repro** — findings without steps or contract ref ## Related - [Findings Communication](../lamina-research-communication/SKILL.md) - [Verify](../lamina-verify/SKILL.md)

Details

Author
aryaniyaps
Repository
aryaniyaps/lamina
Created
3 weeks ago
Last Updated
today
Language
TypeScript
License
Apache-2.0

Similar Skills

Semantically similar based on skill content — not just same category