Use Cases
Real workflows where stato makes the difference.
1. Across Sessions
You spend two sessions tuning QC parameters for mouse cortex scRNA-seq data. Stato captures those parameters, lessons, and decisions so the next session starts with expert knowledge instead of from scratch.
Agent uses default QC thresholds (min_genes=200)
No tissue-specific adjustments
Doesn't know FFPE samples need max_pct_mito=40
Re-discovers mouse uses lowercase mt- prefix
Every session starts from scratch.
Agent reads qc_filtering skill: max_pct_mito=20
Knows cortex retains ~85% of cells at this threshold
FFPE samples: automatically adjusts to 40
Mouse mt- prefix is in lessons_learned
Every session starts with expert knowledge.
# Session 1: agent captures expertise
$ stato init
$ stato crystallize
# Agent writes .stato/ modules from its own knowledge
$ stato bridge --platform claude
✓ Generated CLAUDE.md
# Session 2 (next day, after /compact):
# Agent reads CLAUDE.md automatically
$ stato resume
# Full context restored
2. Across Projects and Teams
A senior researcher captures months of hard-won parameter tuning and lessons into stato modules. A new team member imports those modules and their coding agent starts with senior-level expertise.
New team member's agent uses generic defaults
2 weeks re-learning parameter thresholds
Tribal knowledge scattered across Slack and docs
Senior repeats same explanations
Knowledge transfer takes weeks.
Senior exports expertise as .stato template
New member imports, agent has full skill set
Validated parameters, documented lessons
Agent uses correct thresholds from day one
Knowledge transfer takes one command.
# Senior researcher exports
$ stato snapshot --name "lab-expertise" --template --sanitize
✓ Secrets sanitized in snapshot
✓ Created lab-expertise.stato
# New team member imports
$ stato init
$ stato import lab-expertise.stato
$ stato bridge --platform cursor
✓ Generated .cursorrules
3. Across Platforms
Your team uses Claude Code, Cursor, and Codex across different projects. Stato bridges translate one set of expertise modules into platform-native instruction files for every agent.
Manually sync CLAUDE.md with .cursorrules
Different formats, different conventions
Expertise drifts between platforms
No validation on any platform
Manual copy-paste across platforms.
One .stato/ directory is the single source
Bridge generators handle format differences
All platforms read the same validated expertise
One command regenerates every bridge
One source, every platform.
# Generate all bridge files from one source
$ stato bridge --platform all
✓ Generated CLAUDE.md
✓ Generated .cursorrules
✓ Generated AGENTS.md
✓ Generated README.stato.md
# Each agent reads its native file automatically
4. Web AI to Coding Agent
You spend an hour designing an architecture with Claude.ai. Instead of losing that work when you close the tab, crystallize it into a bundle and import into Claude Code.
Architecture decisions trapped in Claude.ai chat log
Manual copy-paste of key snippets
Coding agent starts with no context
Re-explains design choices every time
Web AI knowledge dies with the tab.
Crystallize web AI conversation into bundle
One command imports all modules
Coding agent has full architectural context
Design decisions preserved in plan.decision_log
Web AI expertise flows into your codebase.
# In your terminal
$ stato crystallize --web --raw | pbcopy
# Paste into Claude.ai / Gemini / ChatGPT
# Save the AI's output as stato_bundle.py
$ stato import-bundle stato_bundle.py
Result: 5 imported, 0 failed
✓ Generated CLAUDE.md
# Open Claude Code, it reads CLAUDE.md automatically