Day-by-day tactical playbook. Start today with what's live (LemmaFit core + lemmafit add), drip content daily, drop shadow later this week as the crescendo. Check off tasks as you go.
lemmafit add — add verification to existing codebases"LemmaFit brings formal verification to your codebase using Dafny to mathematically prove properties..."
Eyes glaze. "Formal verification" = academic, not for me.
"We found a real Next.js SSRF vulnerability — CVE-2024-34351 — before it was public."
"What if your code could prove no bug exists — not just that tests haven't found one yet?"
"Your tests check 10,000 inputs. Formal verification checks all of them. At once."
Start with the consequence. The method becomes interesting after the stakes land.
Real failures a simple invariant catches. You have 70+. Each is a post.
lemmafit add, shadow launch, gallery. Each feature is its own moment.
The Next.js CVE. OSS bugs found. User reactions. Concrete evidence.
Testing vs. proving. What's an invariant? Why simulations lie. Lobste.rs-friendly.
Landing page updates, gallery showcase, docs. The place everything points to.
lemmafit add announcement. Write a short blog post: what it does, who it's for, one example. Draft tonight, publish tomorrow.
ALSO: MAKE SURE THE INSTALL IS ONE COMMAND, WORKS FIRST TRY
lemmafit add blog post: "You can now add formal verification to your existing codebase"
SHORT, PRACTICAL, INCLUDE INSTALL COMMAND + QUICKSTART
lemmafit add FERNANDA
lemmafit add — before/after showing a codebase without verification, then with. Include screenshot/GIF. End with repo link. Also post to Bluesky.
VISUAL PROOF IS KEY — SHOW THE TERMINAL OUTPUT
lemmafit add to the website. Feature section or getting-started update. Make sure the blog post CTA points to the site.
WEBSITE = SOURCE OF TRUTH · BLOG DRIVES TRAFFIC HERE
Days 1-4: Flood the zone with educational posts to own the narrative before you even announce the product. Build an audience that's primed.
Three batches of 5 hand-picked people each. Batch 1 after educational posts land. Batch 2 after lemmafit add. Batch 3 after shadow.
You have 70+ real bugs. Post one per week. Each thread shows: the story, the invariant it broke, why tests missed it, how verification catches it.
Post one "here's how to think about X" post per week. Educational, not salesey. Builds credibility for Week 2 when product makes sense.
Every post points here. Website shows the gallery (10 verified apps), a clear problem statement, and one-click install. This is your MRR driver.
Each tool drop (lemmafit add, shadow) is a moment. Tease before, explain during, celebrate after. Make it an event.
Not everyone uses Claude Code. For people on Cursor, VSCode, Neovim, etc., you need a different path: run verification for them, send results as GIFs, show benefit without friction.
"I'll run shadow on your repo and send you what it finds — no install needed. Sound good?"
Share results + screenshots from non-CC users who tried it. "Caught in 5 minutes."
Build a web app where they paste a GitHub URL and get instant results. No signup.
Post the first thread today. Check tasks as you complete them. This playbook moves fast — you'll have real signal by Day 7. Use that data to adapt.