Launches, kills, operation dispatches, WeeklyG episodes and the monthly numbers, newest first, in one feed.
Football Slam will not continue as a standalone commercial product. The complete, deployed football game successfully served its primary purpose as a fun Fable 5 capability experiment, but no demand, adoption, revenue, or distribution evidence supported continued product investment. Its private source repository and playable Vercel build remain preserved as technical and portfolio learning.
Read the post-mortem →TUI Coasters will not continue as a standalone commercial product. Its substantial terminal simulator and unlisted 3D browser-port companion fulfilled their primary purpose as fun Fable 5 capability experiments, but neither established a packaged release, durable distribution, user adoption, revenue, or a standalone go-to-market thesis. Both repositories remain preserved as companion technical artifacts.
Read the post-mortem →TrailBrake will not continue as a commercial sim-racing coaching product. Its research identified a plausible training gap, but no application shipped and no adoption or revenue evidence was recorded. The owner rejected the student-built-for-students authority model because credible racing instruction and distribution require professional racing or coaching provenance. The repository and research remain preserved as portfolio learning.
Read the post-mortem →FormForge will not deploy as a standalone commercial product. Crowded form-builder economics and weak differentiation could not justify the compliance, integration, infrastructure, and customer-support burden. Its code and research remain preserved as portfolio learning.
Read the post-mortem →The current GSkillPacks commercial product thesis will not continue. Alignment and interrogation workflows became required ceremony that delayed the core value of finding, installing, and using reusable agent skills. The public agentic-skills repository, published skillpacks package, domain, and source history remain preserved as active OSS/shared infrastructure. A lower-friction v2 is a possible revisit path, not an active product commitment.
Read the post-mortem →SnipVault will not continue as a standalone commercial product. Its broad prototype did not establish validated demand or enough differentiation to justify the authentication, AI, database, billing, extension-distribution, security, and support burden. The repository and reusable implementation patterns remain preserved as an archived code asset.
Read the post-mortem →TrackCTL will not continue as a standalone commercial product. General computer-use automation absorbed most of its browserless PostHog setup wedge, while no TrackCTL-specific customer, adoption, release, or revenue evidence established demand for a separate analytics-governance layer. The remaining secrets problem is cross-cutting infrastructure tracked through the EnvBank experiment. TrackCTL's repository, archived implementation, and research remain preserved as portfolio learning.
Read the post-mortem →B4 shipped its planned community-platform scope, but the activation loop failed. Attendance declined, founder/investor pull was not strong enough, and continuing would have required more outreach and fundraising-community energy than the founder could justify.
Read the post-mortem →Payment Provider Blocked
Read the post-mortem →Get every launch, kill and dispatch as it lands.