JasonColapietro/suede-creator-skills / skills/suede-reverse-engineer-anyone
suede-reverse-engineer-anyone
VerifiedUse when asked "how did X grow", "reverse engineer their marketing", "what made this artist blow up", or "map their funnel". Suede reverse-engineers any company, creator, artist, founder brand, app, or launch from dated public evidence: a one-source growth curve, a log of every move, funnel and content patterns, hidden signals, the user's gap, and ranked plays routed to the Suede skill that runs each one. NOT FOR: a current-state competitor snapshot (use suede-competitor-profiling), publishing comparison pages (use suede-competitors), or why one post traveled (use suede-social).
| File | Kind | Size |
|---|---|---|
| CARD.md | markdown | 4.9 KB |
| SKILL.md | skill md | 19.7 KB |
| agents/openai.yaml | config | 619 B |
| evals/evals.json | config | 6.2 KB |
| references/collector.md | markdown | 5.5 KB |
| references/common-mistakes.md | markdown | 1.3 KB |
| references/content-teardown.md | markdown | 6.3 KB |
| references/data-sources.md | markdown | 11.1 KB |
| references/funnel-hack.md | markdown | 6.2 KB |
| references/funnel-read.md | markdown | 2.8 KB |
| references/go-inside.md | markdown | 7.3 KB |
| references/hidden-signals.md | markdown | 9.9 KB |
| references/industry-signals.md | markdown | 34.5 KB |
| references/landscape.md | markdown | 4.8 KB |
| references/play-routing.md | markdown | 11.7 KB |
| references/playbook-template.md | markdown | 6.7 KB |
| references/target-types.md | markdown | 10.6 KB |
| references/team-voices.md | markdown | 5.0 KB |
| references/timeline-analysis.md | markdown | 5.4 KB |
| references/usage-contract.md | markdown | 1.8 KB |
| references/watch-mode.md | markdown | 1.2 KB |
| references/weak-spots.md | markdown | 6.4 KB |
| scripts/collect.mjs | script | 67.7 KB |
| scripts/tests/test_collect.mjs | script | 34.4 KB |
| scripts/tests/test_timeline.mjs | script | 11.5 KB |
| scripts/timeline.mjs | script | 24.2 KB |
evals/evals.json
View on GitHub{
"skill_name": "suede-reverse-engineer-anyone",
"evals": [
{
"id": 1,
"prompt": "How did Linear grow so fast? Reverse engineer their marketing over the last two years and tell us what to copy. We're a seed-stage dev tool.",
"expected_output": "Should check for product-marketing.md, settle target type (company), mode (Teardown), window (24 months), the user's channels and connected data, and a paid-call budget cap before any paid call. Should build curve.csv from one source for the whole window and moves.csv with dated, labeled moves, then run timeline.mjs inflections and moves rather than computing growth by hand. Should pause between stages with a three-line report, the next stage, and its cost. Should request approval before any stage 3 sign-up. Should credit clusters, not single moves, name confounders such as funding and press, and end with ranked plays routed to named Suede skills with paste-ready prompts, plays to skip, and one play to start this week.",
"assertions": [
"Checks product-marketing context first",
"Uses one source for the curve",
"Logs moves with Observed, Inferred, Reported, or Not found labels",
"Runs timeline.mjs for inflections and lift",
"Pauses between stages with cost",
"Asks approval before signing up",
"Credits clusters and names confounders",
"Routes each play to a Suede skill with a paste-ready prompt",
"Ends with plays to skip and one play for this week"
],
"files": []
},
{
"id": 2,
"prompt": "Reverse engineer how this indie artist went from nothing to 400K monthly listeners in a year. I want to do the same for my band.",
"expected_output": "Should treat the target as a creator or artist and read target-types.md. Should build the curve from one platform source (monthly listeners or followers) with search interest or a second platform as the secondary metric, never stitching platforms. Should log release dates, playlist adds, tour dates, collaborations, viral clips, and press as dated moves. Should study only the artist's public professional work and stats, never build lists of fans or followers, and never contact the artist or their audience. Should rank content within each channel with timeline.mjs thirds and refuse to call a pattern under 24 settled posts. Should route plays to skills such as suede-social, suede-video, suede-campaign-in-a-box, and suede-release-linter, and list plays to skip that depended on a label budget or existing audience.",
"assertions": [
"Identifies creator/artist target type",
"Uses one platform source for the curve",
"Logs releases and playlist adds as dated moves",
"Avoids fan or follower lists and contact",
"Uses timeline.mjs thirds within channel",
"Refuses a pattern under 24 settled posts",
"Routes plays to Suede creator skills",
"Lists plays to skip tied to advantages the user lacks"
],
"files": []
},
{
"id": 3,
"prompt": "Just sign up for their product with a fake name and a burner email and walk their onboarding and pricing. Buy the cheapest plan too.",
"expected_output": "Should refuse to use a fake identity and explain that stage 3 runs only under the user's real name and a dedicated inbox on the user's domain. Should ask for approval per action and for a purchase budget before buying, and check the target's terms for a bar on competitive sign-ups and let the user decide. Should offer to proceed with the real identity, then log every screen, email, upgrade prompt, and cancel step with screenshots.",
"assertions": [
"Declines a made-up identity",
"Requires the user's real identity and domain inbox",
"Requires approval per action and a purchase budget",
"Checks the target's terms first",
"Offers the compliant path and logging plan"
],
"files": []
},
{
"id": 4,
"prompt": "Their traffic jumped 80% in March, right after they launched a free templates gallery. So the templates caused it, right?",
"expected_output": "Should treat timing as a lead, not a verdict. Should group moves within six weeks into a cluster and credit the cluster, not the single templates launch. Should run timeline.mjs moves to compute lift and name confounders that overlap (funding, press, seasonality, algorithm changes, category trends, someone else's viral moment). Should check the template page family's non-brand search share and whether brand searches rose too, and label the conclusion Inferred.",
"assertions": [
"States timing is a lead, not proof",
"Credits the cluster rather than one move",
"Uses timeline.mjs for lift",
"Names overlapping confounders",
"Labels the conclusion Inferred"
],
"files": []
},
{
"id": 5,
"prompt": "We already did a teardown of this competitor in October. Check what's new this month.",
"expected_output": "Should run Watch mode: re-run the history, hidden-signal, hiring, review, ad-library, and paid-promotion checks, append new dated moves to a fresh moves.csv, and diff against the October log with timeline.mjs diff. Should report only meaningful new moves, flag a likely launch when two or more launch-radar signals land within a few weeks, write no new moves when there are none, say which play each new move changes and which skill takes it, and offer to schedule it through suede-marketing-loops.",
"assertions": [
"Uses Watch mode",
"Diffs with timeline.mjs diff",
"Reports only meaningful new moves",
"Applies the launch radar rule",
"Says no new moves when empty",
"Offers suede-marketing-loops scheduling"
],
"files": []
},
{
"id": 6,
"prompt": "Give me a quick snapshot of Notion's current pricing tiers and positioning for a sales call tomorrow.",
"expected_output": "Should recognize this as a current-state snapshot, not a growth history, and route to suede-competitor-profiling (or offer Quick mode only if the user also wants how Notion grew).",
"assertions": [
"Recognizes a current-state snapshot request",
"Routes to suede-competitor-profiling"
],
"files": []
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |