一键在 Manus 中运行任何 Skill

invalidate-the-lore

星标0

分支0

更新时间2026年7月2日 09:03

Harvest a team's beliefs about why users behave, convert each into a testable claim, and run the cheapest disproving test — building an append-only lore ledger of confirmed/busted/undecidable verdicts. Use when roadmap arguments run on gut feeling, before any big bet built on an untested assumption, on a quarterly hygiene cadence, or when someone says "users churn because…" without a link attached.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

zakelfassi

zakelfassi/skdd-commons

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Invalidate the Lore

Turn team beliefs into testable claims, test them cheaply, and log the verdicts where nobody can un-log them.

Gate

Works best with gate: instrumented = open, but may run earlier — some lore dies to data that already exists (support tickets, store reviews, git history, server logs). No gate blocks collecting lore; only verdicts need evidence.

Inputs

Beliefs, harvested from: the user/founder directly ("tell me five things you know about your users"), docs and READMEs, PR discussions, support threads, marketing copy (every headline is a belief).
Whatever data exists. GROWTH.md ## Lore ledger as working memory.

Steps

Harvest without judging. Collect beliefs verbatim, attributed and dated. The confident ones and the load-bearing ones first — a belief that steers the roadmap is worth ten idle opinions.
Sharpen each into a claim. Rewrite as something falsifiable: "users churn because onboarding is long" → "churned users abandoned during onboarding at a higher rate than retained users." If a belief can't be sharpened, log it undecidable with the reason — that's a verdict too.
Design the cheapest kill. For each claim, the minimum test: usually a query, sometimes a cohort comparison, occasionally a real experiment (route those to run-the-loop). Order by load-bearing-ness ÷ cost. Independent claims fan out well — one lane per belief, verified separately.
Run and rule. confirmed / busted / undecidable, each with an evidence link. Busted lore gets a one-line epitaph. Never bust by counter-assertion — no evidence, no verdict.
Ledger it. Append to GROWTH.md ## Lore ledger (date, author, belief, status, evidence). Surface the two or three verdicts that should change what the team does next — a ledger nobody reads is a diary.

Conventions

Gut feeling isn't banned; it's promoted — from conclusion to hypothesis. Say this out loud when harvesting; it keeps people honest and unhumiliated.
The ledger is append-only. Corrections are new entries citing old ones.
Treat your own analyses as lore too: a conclusion this skill produced last quarter is a belief this quarter.

Edge Cases

The ledger only ever confirms. Audit the sharpening step — soft claims produce soft confirmations. A healthy ledger busts things.
Belief held by the person paying for the work. Test it exactly as carefully, deliver the verdict exactly as plainly, with the evidence doing the talking.
Two beliefs conflict. Perfect — at most one survives. Test the pair together and log both verdicts.
No data whatsoever. Convert the top beliefs into the instrumentation wishlist and route to instrument-the-truth; log everything harvested as untested so the debt is visible.

同仓库更多 Skills

同仓库

name-the-aha

zakelfassi/skdd-commons

Interrogate a product for its core value — name the observable aha moment, measure time-to-aha, or honestly conclude there is none yet. Use when starting Growth Circle work on any repo, when someone asks "how do we grow this", when GROWTH.md is missing or its Aha section is empty, or before any activation/retention/distribution work is proposed. This is the gatekeeper skill — every other growth skill depends on its output.

2026-07-020

prove-the-value

zakelfassi/skdd-commons

Compile a product's earned, evidence-backed claims from GROWTH.md into a public VALUE.md — the proof-of-value surface that agent-mediated discovery reads — with attestation levels, expiry dates, and honest retractions. Use when creating or refreshing a VALUE.md, when marketing copy needs grounding in evidence, when preparing a product to be evaluated by assistants and their principals' agents, or when a published claim has gone stale.

2026-07-020

run-the-loop

zakelfassi/skdd-commons

Run the growth experiment cadence — propose ranked by keystone impact, pre-register before launch, evaluate by the committed rule, keep/kill without relitigating, log everything. Enforces the virality gate (get-users-to-get-users proposals are parked until activation and retention run green) and the alienation test (auto-kill anything a fully-informed user would resent). Use when the growth cadence comes due (weekly/biweekly), when someone proposes "a growth hack", or when experiment results need an honest verdict.

2026-07-020

forge-the-local-circle

zakelfassi/skdd-commons

Use SkDD to forge project-specialized adaptations of the Growth Circle skills — local skills that carry this product's keystone, event names, query paths, contexts, and cadence baked in, so future sessions (and cheaper models) run the circle without rediscovering it. Use when the generic pack has run against a project a few times and keeps re-deriving the same context, when handing growth work to scheduled or lighter-weight agents, or when a project's circle has stabilized enough to be worth freezing as skills.

2026-07-020

wind-the-clock

zakelfassi/skdd-commons

Make the growth circle self-running — schedule its recurring checks (loop cadence, keystone sunset, first-light freshness, AEO probes, trust-ledger review) using whatever scheduling primitive the current harness offers, and record every schedule in GROWTH.md so it survives harness changes. Use when the circle's gates are open but nothing recurs, when a keystone sunset or instrumentation check exists only as a date nobody watches, after run-the-loop establishes a cadence, or when growth work keeps happening "whenever someone remembers".

2026-07-020

hire-the-circle

zakelfassi/skdd-commons

Turn growth-team values into working artifacts — role scorecards, interview kits with listen-fors — and map which rings of the growth circle run on agents versus humans, with explicit approval boundaries. Use when hiring a first growth person, when staffing growth work across a human/agent team, when a growth hire is failing and nobody can say why, or when deciding what an autonomous agent may and may not do to users.

2026-07-020

name	invalidate-the-lore
description	Harvest a team's beliefs about why users behave, convert each into a testable claim, and run the cheapest disproving test — building an append-only lore ledger of confirmed/busted/undecidable verdicts. Use when roadmap arguments run on gut feeling, before any big bet built on an untested assumption, on a quarterly hygiene cadence, or when someone says "users churn because…" without a link attached.
metadata	{"pack":"2026-07-growth-circle","forged-by":"claude-fable-5","forged-from":"session-2026-07-02-growth-circle — the Growth Circle protocol (github.com/zakelfassi/growth-circle)","forged-reason":"Every team accumulates folklore about its users, usually asserted by whoever struts most confidently. The cultural unlock isn't better guessing — it's the first time a favorite belief dies to a twenty-line query, after which assertions start arriving with evidence attached.","usage-count":"0"}

Invalidate the Lore

Turn team beliefs into testable claims, test them cheaply, and log the verdicts where nobody can un-log them.

Gate

Inputs

Beliefs, harvested from: the user/founder directly ("tell me five things you know about your users"), docs and READMEs, PR discussions, support threads, marketing copy (every headline is a belief).
Whatever data exists. GROWTH.md ## Lore ledger as working memory.

Steps

Harvest without judging. Collect beliefs verbatim, attributed and dated. The confident ones and the load-bearing ones first — a belief that steers the roadmap is worth ten idle opinions.
Sharpen each into a claim. Rewrite as something falsifiable: "users churn because onboarding is long" → "churned users abandoned during onboarding at a higher rate than retained users." If a belief can't be sharpened, log it undecidable with the reason — that's a verdict too.
Design the cheapest kill. For each claim, the minimum test: usually a query, sometimes a cohort comparison, occasionally a real experiment (route those to run-the-loop). Order by load-bearing-ness ÷ cost. Independent claims fan out well — one lane per belief, verified separately.
Run and rule. confirmed / busted / undecidable, each with an evidence link. Busted lore gets a one-line epitaph. Never bust by counter-assertion — no evidence, no verdict.
Ledger it. Append to GROWTH.md ## Lore ledger (date, author, belief, status, evidence). Surface the two or three verdicts that should change what the team does next — a ledger nobody reads is a diary.

Conventions

Gut feeling isn't banned; it's promoted — from conclusion to hypothesis. Say this out loud when harvesting; it keeps people honest and unhumiliated.
The ledger is append-only. Corrections are new entries citing old ones.
Treat your own analyses as lore too: a conclusion this skill produced last quarter is a belief this quarter.

Edge Cases

The ledger only ever confirms. Audit the sharpening step — soft claims produce soft confirmations. A healthy ledger busts things.
Belief held by the person paying for the work. Test it exactly as carefully, deliver the verdict exactly as plainly, with the evidence doing the talking.
Two beliefs conflict. Perfect — at most one survives. Test the pair together and log both verdicts.
No data whatsoever. Convert the top beliefs into the instrumentation wishlist and route to instrument-the-truth; log everything harvested as untested so the debt is visible.