--- name: parked-mistify-salvage-audit description: "The mistify-master salvage audit — 'what is RobCo MISSING that the old site had'. It ran; the ONE salvageable thing was the DECODE effect, everything else was rejected as not-a-fit. Do NOT re-run without the owner's explicit go. Keeps the read-only constraint on that repo and the port-the-code vs steal-the-idea framing." metadata: node_type: memory type: project originSessionId: 0b7c5f41-bb34-4ab4-93be-fdb363f5eb40 --- ## THE VERDICT Owner, 2026-07-14: *"we did run the what is RobCo missing thing, that's how we got the decode effect."* The salvage audit of `C:\Dev\mistify-master` (his old site) was performed. **Its sole surviving output was the DECODE effect** — text scrambling out of block-glyph static. **Everything else in the old site was rejected as not-a-fit for RobCo.** **Why the rest was rejected:** the site is ~6 years old and framework/bundler-based. RobCo is plain global-scope JS with no build step, offline-first, must-stay-free, mobile-primary and deterministic-native. Most of what was there either depended on that framework, assumed paid or always-online infra, or was a desktop-shaped interaction that doesn't survive contact with a phone. The ideas that *were* good were mostly already covered. ⚠ **Do NOT re-run this on your own initiative.** But note the file's own closing instruction, which outranks any header: **wait for the owner's explicit go** before touching that repo again for any follow-up pass. *Corrected 2026-07-20: this file previously self-contradicted — a "CLOSED, don't re-run" header against a closing "wait for the owner's explicit go". Resolved in favour of the owner instruction: it ran, and any further pass still needs his go.* ## ⚠ THE REPO IS STRICTLY READ-ONLY `C:\Dev\mistify-master` is a site the owner started ~6 years ago, stopped working on ~3 years ago, and has had other Claude sessions actively fixing. **No writes, no `git` operations, no branch switching, no commits.** Another session may be live in that tree; touching it would collide. ## THE FRAMING THAT MADE THE AUDIT USEFUL (reusable for any salvage pass) The owner was explicit that an inventory of the `components/` folder was the FLOOR, not the ceiling: > *"the components file is features that have been separated and organized for use if wanted. **I want more than that though, I want the code to be evaluated and figure out what RobCo is MISSING that the old site had** (that fits for RobCo)."* So the deliverable is a **capability-gap analysis**, not a list of portable snippets. **Separate these two and never conflate them:** - **"PORT THE CODE"** — a self-contained function, CSS technique, algorithm or asset. Drop-in or near-drop-in. - **"STEAL THE IDEA"** — the concept is strong, the implementation doesn't fit RobCo's architecture, and it would need rebuilding RobCo-style. Still worth knowing; just priced honestly. **Per finding:** what it is · cosmetic or functional · **why it fits ROBCO specifically** (not merely "this is neat") · rough cost to bring across. Triaged, so the owner can skim the top and ignore the rest. **Read `library/BRAIN_DUMP.md` FIRST** so any such audit knows RobCo's hard constraints before judging fit. ⭐ **Be honest if the answer is "not much."** An audit that manufactures twelve findings to justify itself is worse than one that surfaces two real ones. (Same standard now applied to the independent code audits — see [[engineering-metrics-log]].) This audit returning exactly one keeper is the standard working, not the standard failing. ## ADVANTAGE WORTH REMEMBERING It is the owner's OWN prior work: **no licensing question, no unfamiliar third-party code, and he already knows why he built each piece.** Salvage value from his own archives is unusually cheap compared to inventing a feature from scratch — worth remembering if another old project of his ever comes up. Related: [[roadmap-parked-placement]], [[docked-ideas]], [[player-control-principle]].