EXPERIMENTS · VALIDATION BEFORE STATUS
Things I build to find out.
I use experiments to test problems, mechanisms, workflows, and demand without presenting unfinished work as mature software. Each track states what exists now and what still needs evidence.
Current validation and utility work
Portability Check
Fast validation
Assessment and remediation method for AI-built application portability. The method exists; outside assessment demand, automated scanning, and paid remediation still need proof.
Open →MindVergent™
Exploration track
Professional reputation through work other people can test, improve, and verify. Founding intake and research surfaces can be tested now; formal contribution scoring, matching, and commerce remain roadmap.
Context →Assemblerer
Supporting tool
Inspectable company-manifest authoring surface. Useful supporting infrastructure around bounded AI operating context rather than a separate top-level venture claim.
Open →QuotaWatch
Published utility
Local-first usage and quota tracking. A useful shipped utility and product proof, but not currently treated as a core venture.
Open →Playlist Surgeon
Open-source Windows utility
Bulk artist/track cleanup for editable Spotify playlists. MIT-licensed local EXE source with PKCE, blocklist rules, backups, and an explicit compatibility policy to follow Spotify Web API changes.
Open →Right-Click Menu Manager
Published utility
Local-first Windows context-menu control. A concrete utility with backup and rollback rather than a broader company thesis.
Open →Shared Skills
Public methods
Reusable operating rules and specifications. Skills capture methods that can support products, client work, or future experiments without being marketed as standalone companies.
Open →Promotion path
PROBLEM → EXPERIMENT → TOOL → PRODUCT
I give an experiment stronger status when the problem repeats, the solution remains useful outside its original context, and outside evidence supports continued investment. Newness alone is not evidence.
Why I keep the separation.
I can build quickly without presenting every prototype as a new SaaS company. Products remain credible, experiments remain honest, and useful one-off tools can stay useful without being forced into a venture narrative.