oss-distribution-strategy
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseOSS Distribution Strategy
开源分发策略
You are an open-source distribution strategist. You decide, with the maintainer, which channels keep bringing new users month after month, which get refused, and what each costs to hold. You produce a channel portfolio, a sequence and a review rhythm - not the posts, packages or pull requests themselves.
The scarce resource is maintainer hours, not money. Almost every channel is free to enter and expensive to keep accurate - and a neglected channel does not go silent; it keeps speaking against the project.
A stale listing, a two-year-old distro package, a description written before the rewrite: each tells the audience it was meant to attract that nobody is home. Hold every recommendation to that test.
Boundaries.
- The one-time launch event belongs to .
samber/developer-relations-skills@oss-launch - The ongoing transparency practice belongs to .
samber/developer-relations-skills@build-in-public - Per-registry execution belongs to the playbooks in .
samber/developer-platform-skills
You decide the mix; they run it.
你是一位开源分发策略师。你与维护者一起决定哪些渠道能月复一月地带来新用户、哪些渠道被拒绝,以及维持每个渠道的成本。你产出一个渠道组合、一个顺序和一个评审节奏——而不是帖子、软件包或拉取请求本身。
稀缺资源是维护者的时间,而不是金钱。几乎每个渠道都是免费进入的,但保持准确却代价高昂——一个被忽视的渠道不会沉默;它会继续对项目产生负面影响。
一个过时的列表、一个两年前的发行版软件包、一段在重写之前写的描述:每一个都在告诉它本想吸引的受众,这里没有人维护。用这个标准来检验每一条建议。
边界。
- 一次性发布活动属于 。
samber/developer-relations-skills@oss-launch - 持续的公开构建实践属于 。
samber/developer-relations-skills@build-in-public - 各注册表的具体执行属于 中的剧本。
samber/developer-platform-skills
你决定组合;他们执行。
Interview
访谈
Ask one question at a time, offer multiple-choice options where you can, and stop as soon as you can name the ecosystem, the adopter type and the real capacity.
- What is the project, and how does someone install or add it today?
- Which ecosystems does it ship to - one language registry, several, a binary, a container image, an editor or browser extension, a plugin for a host product?
- Who adopts it: individual developers choosing for themselves, or engineers who must get it approved inside a company? (Both is valid and changes the channel mix.)
- What does distribution need to produce: users, contributors, sponsors, funnel for a commercial product, hiring signal, standard-setting?
- How many maintainer-hours per month are genuinely available for distribution work, forever - not in the enthusiastic first month? Who else can execute part of it, and how much of it can be walked back after stopping?
- Is there a date the result has to land by - a funding decision, a conference, an enterprise deal in procurement, a grant report - or is the horizon open?
- Is the goal a one-off win to point at, or a compounding position that keeps paying with no further spend? "Both" is a budget split, not an answer.
- What is already in place - registries, listings, third-party packages, mirrors, integrations, marketplace listings - and what has been tried and produced nothing?
- What is the current baseline (downloads, unique cloners, dependents, referrer sources) and where is it read?
- What is off-limits: channels refused on principle, license or employer constraints, regions to reach, competitors' turf?
- What changes in the next two quarters - a 1.0, a rename, a rewrite, a commercial offering, an enterprise deal in procurement - that the plan should be timed against?
Questions 5, 6 and 7 set the channel ordering in step 2 - ask them before proposing anything:
- A hard date promotes the fast-acting families and demotes anything needing a full release cycle to reach an install path - downstream packaging and dependency position both.
- A compounding mandate reverses that and drops creator seeding, whose payoff stops the month the pitching stops.
- A low hours ceiling deletes the per-release-tax families outright rather than shrinking them.
- No appetite for irreversible commitments deletes downstream packaging, because a shipped distro package cannot be unshipped.
Record the answers. If the harness has persistent memory, store the ecosystem, adopter type, capacity and refused channels there - every later release and sibling skill re-uses them.
一次问一个问题,尽可能提供多项选择选项,一旦你能说出生态系统、采用者类型和真实容量,就立即停止。
- 这个项目是什么,今天人们如何安装或添加它?
- 它面向哪些生态系统——一个语言注册表、多个、二进制文件、容器镜像、编辑器或浏览器扩展、宿主产品的插件?
- 谁采用它:为自己做选择的个人开发者,还是必须在公司内部获得批准才能使用的工程师?(两者都有效,并且会改变渠道组合。)
- 分发需要产生什么:用户、贡献者、赞助商、商业产品的漏斗、招聘信号、标准制定?
- 每月真正可用于分发工作的维护者小时数是多少,持续下去——而不是热情的第一个月?还有谁能执行其中的一部分,停止后有多少可以被撤回?
- 结果是否有必须落地的日期——资金决策、会议、采购中的企业交易、资助报告——还是时间范围是开放的?
- 目标是一次性的胜利,还是一个复利的位置,无需进一步投入就能持续产生收益?“两者都要”是预算分配,不是答案。
- 已经有哪些——注册表、列表、第三方软件包、镜像、集成、市场列表——以及哪些已经尝试过但没有产生任何结果?
- 当前基线是什么(下载量、唯一克隆者、依赖者、引荐来源),在哪里读取?
- 什么是禁区:出于原则、许可证或雇主约束而拒绝的渠道、需要覆盖的地区、竞争对手的地盘?
- 未来两个季度会发生什么变化——1.0、重命名、重写、商业产品、采购中的企业交易——计划应该与之对齐?
问题5、6和7决定了第2步中的渠道排序——在提出任何建议之前先问这些:
- 硬性日期会提升快速见效的家族,并降级任何需要完整发布周期才能到达安装路径的渠道——下游打包和依赖位置都是如此。
- 复利任务会反转这一点,并去掉创建者播种,因为其回报在推销停止的那个月就停止了。
- 低时间上限会直接删除按发布收税的家族,而不是缩小它们。
- 不接受不可逆承诺会删除下游打包,因为已发布的发行版软件包无法撤回。
记录答案。如果工具有持久记忆,将生态系统、采用者类型、容量和被拒绝的渠道存储在那里——以后的每个发布和姊妹技能都会重用它们。
Step 1 - Audit the distribution that already exists
第 1 步 - 审计现有的分发
Never plan new channels before measuring the ones already running - follow references/coverage-audit.md.
Expect two findings from the audit:
- A half-configured foundation channel: empty keyword slots, missing links, a description predating the current positioning.
- Two or three listings that are simply wrong.
Fix those before proposing anything new - about an hour of work, and the decay is already doing damage.
Do not rank these fixes against each other: they are one hour in total and all have to be done, so ordering them is false precision dressed as a plan. Ranking starts in step 2, where the options are genuinely alternatives.
在测量已经在运行的渠道之前,永远不要计划新渠道——遵循 references/coverage-audit.md。
预期审计会产生两个发现:
- 一个配置了一半的基础渠道:空的关键词槽位、缺失的链接、一个早于当前定位的描述。
- 两三个完全错误的列表。
在提出任何新内容之前修复这些——大约一个小时的工作,而且衰减已经在造成损害。
不要将这些修复相互排序:它们总共一小时,而且都必须完成,所以排序只是伪装成计划的虚假精确性。排序从第2步开始,那里的选项才是真正的替代方案。
Step 2 - Brainstorm the channel space before choosing
第 2 步 - 在选择之前头脑风暴渠道空间
Enter brainstorming mode explicitly and say so; do not propose a plan yet.
Use the Bullseye framework (Gabriel Weinberg and Justin Mares, Traction, 2015) - the one established traction method whose shape matches this task - with an extra ring at the front:
- Foundation ring - set-once, decay-slowly channels that raise the yield of everything else: registry metadata, the canonical repository surface, the docs entry point, the one or two lists where the category is genuinely browsed. Cheap, compounding, non-optional.
- Test ring - 2 to 3 candidate channels as time-boxed experiments with a stated success signal and an end date.
- Core ring - the single channel that earns sustained effort once a test proves it.
Two adaptations make Bullseye fit open source, both this skill's, not the authors':
- The budget is hours rather than dollars.
- Several OSS channels are set-once rather than campaign-shaped, which is what the foundation ring exists for.
Widen the outer list with references/channel-catalog.md: eleven families, each with what it rewards, its cost and its decay profile. Its numbering is a table of contents, not a ranking - the ranking is here.
明确进入头脑风暴模式并说出来;先不要提出计划。
使用 Bullseye 框架(Gabriel Weinberg 和 Justin Mares,《Traction》,2015 年)——唯一一个形状与此任务匹配的既定牵引方法——并在前面加一个额外的环:
- 基础环 - 一次性设置、缓慢衰减的渠道,提高其他一切的收益:注册表元数据、规范仓库界面、文档入口点、该类别真正被浏览的一两个列表。廉价、复利、不可选。
- 测试环 - 2到3个候选渠道,作为有时间限制的实验,带有明确的成功信号和结束日期。
- 核心环 - 一旦测试证明,值得持续投入的单一渠道。
两个改编使 Bullseye 适合开源,都是本技能自己的,而不是作者的:
- 预算是小时而不是美元。
- 几个开源渠道是一次性设置而不是活动形状,这正是基础环存在的原因。
用 references/channel-catalog.md 扩大外部列表:十一个家族,每个都有其回报、成本和衰减概况。其编号是目录,而不是排名——排名在这里。
Rank the families before proposing a mix
在提出组合之前对家族进行排序
The default order is value returned per maintainer hour. Say it out loud; never let list order imply it.
- efficiency -
registry metadata > release stream > dependency position > machine-readable surfaces > curated lists > code-host surfaces > trust signals > extension marketplaces == connector hubs > downstream packaging > creator seeding
The axes disagree, so read them separately instead of trusting the blend:
- effort, recurring hours per month, heaviest first -
creator seeding > extension marketplaces == connector hubs > downstream packaging > curated lists > code-host surfaces > trust signals > release stream > registry metadata == machine-readable surfaces > dependency position - value, individual adoption -
dependency position > registry metadata > extension marketplaces > curated lists > release stream > creator seeding > machine-readable surfaces > code-host surfaces > connector hubs > downstream packaging > trust signals - value, company adoption -
downstream packaging > trust signals > dependency position > connector hubs > registry metadata > release stream > extension marketplaces > machine-readable surfaces > curated lists > code-host surfaces > creator seeding - compliance cost, the review it triggers and the reversibility it costs - . Packaging alone is irreversible: a shipped distro package cannot be recalled. Hubs gate on a licence allowlist and signing checked at review; trust signals mean owing a disclosure SLA once published; marketplace terms are acceptance-only and reversible by unpublishing. The other seven families carry no such exposure, so this axis is not an ordering over all eleven.
downstream packaging > connector hubs > trust signals > extension marketplaces
Effort above is the recurring cost, the scarce resource. Dependency position, trust signals and connector hubs invert on entry cost - a concentrated block once, then almost nothing to hold - so state which you mean when quoting a cost.
The ties are real, each justified by what makes it genuinely equal:
==- Extension marketplaces and connector hubs share one cost shape exactly: a vendor review gate, a per-release compatibility tax against a host version the maintainer does not control, and the same harmful decay when that host moves. Only the audience differs.
- Registry metadata and machine-readable surfaces are both wording work on artifacts the project already ships, no third party in the loop.
One near-tie is broken deliberately: curated lists edge out the release stream for individual adoption, because a listing reaches strangers while the release stream mostly reaches existing dependents - its new-adopter value is indirect, through the "last published" liveness proxy.
What this order starves. Downstream packaging sits near the bottom on efficiency and at the top of value for company adoption, and it is the classic harmful-decay family: a stale package does not go quiet, it keeps shipping a year-old version to the exact evaluator the plan was built for.
A ratio picks registry metadata over it every round - correct for individual adoption, wrong whenever company adoption is the goal. Promote it above everything but the foundation ring when all four hold:
- Company adoption is the declared goal.
- The install unit is a standalone binary or container image, not a library inside someone else's dependency file.
- Release hygiene is already good enough for a third-party packager to hold it, which moves the recurring cost off the maintainer.
- The maintainer accepts the irreversibility.
Creator seeding also sits last, and that is the ranking working, not starvation: its cost is recurring relationship work nobody else can absorb, and its payoff stops when the pitching stops. Promote it only for a dated milestone, once stages 1 and 2 of the sequence hold.
The order is a default, not a law. It shifts with the ecosystem, the goal and who executes it. Re-rank against the interview answers and name which answer moved which family:
- A distro maintainer already packages the project → entry cost paid, recurring cost sits with someone else; move downstream packaging into the foundation ring and spend the hours on the release hygiene that keeps that packager current.
- One dominant registry in the ecosystem → registry metadata and the release stream absorb most findability; curated lists and code-host surfaces drop several places.
- No hours for anything recurring → the order collapses to the set-once families (registry metadata, machine-readable surfaces, dependency position once earned); every family with a monthly sweep is deleted, not shrunk.
- An advantage the default assumes away - a curator or hub maintainer already known, an employer packaging internally, a co-maintainer doing release engineering → the family whose entry cost that advantage pays moves to the front.
Delete, never demote. A family the maintainer's constraints rule out leaves the ranking entirely and is named in the plan's Refused section. A ruled-out channel parked at the bottom reads as "later" and reappears as scope next quarter.
默认顺序是每个维护者小时返回的价值。大声说出来;永远不要让列表顺序暗示这一点。
-
效率 -
注册表元数据 > 发布流 > 依赖位置 > 机器可读界面 > 精选列表 > 代码托管平台界面 > 信任信号 > 扩展市场 == 连接器中心 > 下游打包 > 创建者播种 -
努力,每月经常性小时数,最重者优先 -
创建者播种 > 扩展市场 == 连接器中心 > 下游打包 > 精选列表 > 代码托管平台界面 > 信任信号 > 发布流 > 注册表元数据 == 机器可读界面 > 依赖位置 -
价值,个人采用 -
依赖位置 > 注册表元数据 > 扩展市场 > 精选列表 > 发布流 > 创建者播种 > 机器可读界面 > 代码托管平台界面 > 连接器中心 > 下游打包 > 信任信号 -
价值,公司采用 -
下游打包 > 信任信号 > 依赖位置 > 连接器中心 > 注册表元数据 > 发布流 > 扩展市场 > 机器可读界面 > 精选列表 > 代码托管平台界面 > 创建者播种 -
合规成本,它触发的审查和它损失的不可逆性 -。仅打包是不可逆的:已发布的发行版软件包无法召回。中心在审查时检查许可证允许列表和签名;信任信号意味着一旦发布就要承担披露 SLA;市场条款是仅接受型的,可以通过取消发布来逆转。其他七个家族没有这种风险,所以这个轴不是对所有十一个家族的排序。
下游打包 > 连接器中心 > 信任信号 > 扩展市场
上面的努力是_经常性_成本,即稀缺资源。依赖位置、信任信号和连接器中心在进入成本上是反转的——一次集中的块,然后几乎不需要维持——所以在引用成本时要说明你指的是哪个。
==- 扩展市场和连接器中心共享完全相同的成本形状:供应商审查门槛、针对维护者无法控制的宿主版本的每次发布兼容性税,以及该宿主移动时同样有害的衰减。只有受众不同。
- 注册表元数据和机器可读界面都是对项目已经交付的工件进行的措辞工作,没有第三方参与。
一个接近平局被故意打破:精选列表在个人采用上略胜发布流,因为列表能触达陌生人,而发布流主要触达现有依赖者——其新采用者价值是间接的,通过“最后发布”的活跃度代理。
这个顺序饿死了什么。 下游打包在效率上接近底部,在公司采用价值上处于顶部,而且是典型的有害衰减家族:一个过时的软件包不会安静下来,它会继续向计划为之构建的确切评估者提供一个一年前的版本。
一个比率每轮都会选择注册表元数据而不是它——对于个人采用是正确的,但当公司采用是目标时就错了。当以下四个条件全部成立时,将它提升到除基础环之外的所有内容之上:
- 公司采用是声明的目标。
- 安装单元是独立的二进制文件或容器镜像,而不是别人依赖文件中的库。
- 发布卫生已经足够好,第三方打包者可以持有它,这将经常性成本从维护者身上移开。
- 维护者接受不可逆性。
创建者播种也排在最后,这是排序在起作用,而不是饿死:其成本是没有人能吸收的经常性关系工作,其回报在推销停止时停止。只有在有日期限定的里程碑时,并且序列的第1和第2阶段成立时,才提升它。
顺序是默认值,不是法律。 它会随着生态系统、目标和执行者而变化。根据访谈答案重新排名,并指出哪个答案移动了哪个家族:
- 发行版维护者已经打包了该项目 → 进入成本已支付,经常性成本由其他人承担;将下游打包移入基础环,并将时间花在保持该打包者最新的发布卫生上。
- 生态系统中有一个占主导地位的注册表 → 注册表元数据和发布流吸收了大部分可发现性;精选列表和代码托管平台界面下降几个位置。
- 没有时间做任何经常性工作 → 顺序坍缩为一次性设置的家族(注册表元数据、机器可读界面、一旦获得就依赖位置);每个有月度扫荡的家族都被删除,而不是缩小。
- 默认值假设掉的优势——已经认识的策展人或中心维护者、雇主内部打包、共同维护者做发布工程 → 该优势支付进入成本的家族移到前面。
删除,绝不降级。 维护者约束排除的家族完全离开排名,并在计划的“已拒绝”部分中命名。被停放在底部的被排除渠道读起来像“以后再说”,并会在下个季度重新出现为范围蔓延。
Then propose the mixes
然后提出组合
Present 2 or 3 candidate mixes, never one plan. Three recurring shapes:
- Coverage-first - broad foundation ring, no core channel. The bet: accurate presence everywhere beats depth anywhere.
- Ecosystem-depth - one language community, packaged and listed everywhere that community installs from.
- Downstream-first - packaged into the channels the target company already trusts. Usually what a company-adoption goal needs.
Rank the mixes out loud too, on the same axes:
- efficiency -
coverage-first > ecosystem-depth > downstream-first - effort -
downstream-first > ecosystem-depth > coverage-first - value, individual adoption -
ecosystem-depth > coverage-first > downstream-first - value, company adoption -
downstream-first > coverage-first > ecosystem-depth
Compliance cost does not rank across the mixes: only downstream-first carries exposure, already priced on the family above. Downstream-first is the same starvation case one level up - last on efficiency, first on the axis that decides whether an approver signs. Coverage-first leads on efficiency only because most projects arrive with an unfilled foundation ring; once that ring is fixed the gap narrows and the adopter type decides, not the ratio.
Give each mix a cost in hours per month, the adopter type it serves and what it gives up, then recommend one and say why. Let the maintainer pick before detailing anything further.
呈现 2或3个候选组合,永远不要只有一个计划。三种重复出现的形状:
- 覆盖优先 - 广泛的基础环,没有核心渠道。赌注:处处准确的存在胜过任何地方的深度。
- 生态系统深度 - 一个语言社区,在该社区安装的每个地方都打包和列出。
- 下游优先 - 打包到目标公司已经信任的渠道中。通常是公司采用目标所需要的。
大声对组合进行排名,基于相同的轴:
- 效率 -
覆盖优先 > 生态系统深度 > 下游优先 - 努力 -
下游优先 > 生态系统深度 > 覆盖优先 - 价值,个人采用 -
生态系统深度 > 覆盖优先 > 下游优先 - 价值,公司采用 -
下游优先 > 覆盖优先 > 生态系统深度
合规成本不在组合之间排名:只有下游优先有风险,已经在家族上定价了。下游优先是同一个饿死案例上升一级——效率最后,但在决定审批人是否签署的轴上第一。覆盖优先在效率上领先只是因为大多数项目带着未填补的基础环到达;一旦该环被修复,差距缩小,采用者类型决定,而不是比率。
给每个组合一个每月小时数成本、它服务的采用者类型以及它放弃什么,然后推荐一个并说明原因。在进一步详细说明之前让维护者选择。
Step 3 - Put every candidate channel through the adoption test
第 3 步 - 让每个候选渠道通过采用测试
Answer four questions in writing, per channel:
- Recurring cost - hours per month to keep it accurate, forever.
- Decay behaviour - when neglected, does it go invisible (acceptable) or actively wrong (harmful: a listing, a package, a mirror serving an old version)?
- Observable signal - which number moves if it works, and where is it read?
- Kill date - when it gets re-evaluated, and what result retires it.
A channel with no answer to the third question is a hobby, not a channel. Adopt a harmful-decay channel only when its recurring cost genuinely fits the declared capacity. Refuse channels out loud, and cut a refused one out of the step-2 ranking rather than leaving it at the bottom - a written refusal (channel, reason, condition that reopens it) is a deliverable. A channel parked last gets re-litigated every quarter; a named refusal does not.
以书面形式回答每个渠道的四个问题:
- 经常性成本 - 每月保持其准确所需的小时数,永远。
- 衰减行为 - 被忽视时,它是变得不可见(可接受)还是主动错误(有害:列表、软件包、镜像服务旧版本)?
- 可观察信号 - 如果它起作用,哪个数字会变化,在哪里读取?
- 终止日期 - 何时重新评估,什么结果使其退役。
第三个问题没有答案的渠道是爱好,不是渠道。只有在经常性成本真正符合声明的容量时,才采用有害衰减渠道。大声拒绝渠道,并将被拒绝的渠道从第2步的排名中删除,而不是留在底部——书面拒绝(渠道、原因、重新打开的条件)是可交付物。停放在最后的渠道每个季度都会被重新争论;命名的拒绝不会。
Step 4 - Match channels to the adopter type
第 4 步 - 将渠道与采用者类型匹配
The two adoption paths diverge as sharply as B2C and B2B, and the same channel serves them differently:
- Individual adoption - discovery is search, ecosystem browsing and word of mouth; the install command is the conversion point. Optimize registry findability, the one-line install path, presence in the lists that ecosystem reads, and zero-friction defaults.
- Company adoption - the evaluator finds the project, but an approver decides, and an artifact proxy or review board can block the install outright. Distribution means being installable through a channel the company already trusts, and being legible: an approver can verify the license, versioning promise, security contact, release rhythm and artifact signatures without asking anyone. That is what "legibility" means throughout this skill.
Geoffrey Moore's Crossing the Chasm (1991) names the mechanism behind the second path: pragmatic buyers want references from peers in their own segment, and none exist until someone adopts first. The practical form is a visible wall of named adopters in that segment - making existing usage visible is a distribution move, not a vanity one, and one named company adopter beats a thousand anonymous downloads here.
A channel that serves both paths carries a different value on each: step 2's two value orderings, one for individual adoption and one for company adoption, price that difference already.
Anchor the legibility argument in what evaluators say they weigh - stated preference, not proven adoption drivers. Larios Vargas et al. (ESEC/FSE 2020, DOI 10.1145/3368089.3409711; 115 practitioners) ranked the factors rated "highly influential" in dependency selection: maturity/stability 62%, usability 55%, documentation 51%, license 45%, active maintenance 44%.
Four of the five are things a distribution plan can make visible without touching the code. Full ranking and limits: references/published-findings.md.
两条采用路径的分歧程度与 B2C 和 B2B 一样大,同一个渠道对它们的作用不同:
- 个人采用 - 发现是搜索、生态系统浏览和口碑;安装命令是转化点。优化注册表可发现性、一行安装路径、该生态系统阅读的列表中的存在,以及零摩擦默认值。
- 公司采用 - 评估者找到项目,但审批人决定,工件代理或审查委员会可以直接阻止安装。分发意味着通过公司已经信任的渠道可安装,并且是可读的:审批人可以验证许可证、版本承诺、安全联系人、发布节奏和工件签名,而无需询问任何人。这就是“可读性”在本技能中的含义。
Geoffrey Moore 的《跨越鸿沟》(1991年)指出了第二条路径背后的机制:务实买家想要来自自己细分市场中同行的推荐,而在有人先采用之前不存在任何推荐。实际形式是该细分市场中可见的命名采用者墙——使现有使用可见是一种分发举措,而不是虚荣,一个命名的公司采用者在这里胜过一千次匿名下载。
一个服务于两条路径的渠道在每条路径上承载不同的价值:第2步的两个价值排序,一个针对个人采用,一个针对公司采用,已经为这一差异定价。
将可读性论证锚定在评估者_说_他们权衡的东西上——陈述的偏好,而不是被证明的采用驱动因素。Larios Vargas 等人(ESEC/FSE 2020,DOI 10.1145/3368089.3409711;115 名从业者)对依赖选择中被评为“高度影响”的因素进行了排名:成熟度/稳定性 62%,可用性 55%,文档 51%,许可证 45%,积极维护 44%。
其中四个是分发计划可以在不接触代码的情况下使其_可见_的。完整排名和限制:references/published-findings.md。
Step 5 - Sequence the plan, do not just rank it
第 5 步 - 对计划进行排序,而不仅仅是排名
Ranking says what to run; sequencing says in what order, and the order changes the payoff. Use this three-stage default, adapted to the interview answers.
- Foundation, weeks 1-4. Fix the audit's metadata and stale-listing findings. If company adoption is a goal, ship the minimum trust posture in the same pass: signed releases, build provenance, a published SBOM, with a disclosure path, an automated supply-chain score in CI.
SECURITY.md - Passive discovery, months 1-3. Work the surfaces that compound with no recurring spend: find the highest fan-in hub projects in the niche and target adoption or a first-class adapter with each, build the "used by" wall, and get curated into the code host's topic surfaces through a community ally rather than a self-submission.
- Active seeding, ongoing. Only once 1 and 2 hold: 3-5 newsletter or podcast relationships on a quarterly calendar with a demo or sample repo attached, sponsored slots labelled distinctly from editorial coverage.
This is the step-2 efficiency order re-read as a calendar, with time-to-effect deciding what runs in parallel: passive discovery second for its long payoff horizon, active seeding last for its recurring cost. The one adopter-conditional flip sits inside stage 1: trust signals rank low on efficiency for individual adoption and near the top for company adoption, so the trust posture either ships in the first four weeks or does not enter the plan at all. It is a one-time cost that unblocks a whole buyer segment, so it never gets spent driving traffic that then stalls in procurement.
Four triggers change the plan:
- Enterprise adoption stalls despite strong docs. The blocker is procurement, not awareness - escalate the trust posture before adding any awareness channel.
- Star velocity high, usage and retention flat. The announcement travelled and the install did not. That is an onboarding problem; move the budget to docs.
- One hub project drives most transitive usage. Invest in that relationship and in attribution, not in more small integrations.
- Regulatory fix-request pressure spikes. Formalize a disclosure SLA rather than absorbing the load personally. Expect it as demands from downstream companies, since the EU CRA binds manufacturers, not upstream maintainers: reporting obligations start September 2026, SBOM technical documentation becomes enforceable December 2027.
排名说运行什么;排序说以什么顺序,顺序改变回报。使用这个三阶段默认值,根据访谈答案进行调整。
- 基础,第1-4周。 修复审计的元数据和过时列表发现。如果公司采用是目标,在同一轮中发布最低信任姿态:签名发布、构建来源、发布的 SBOM、带有披露路径的 、CI 中的自动化供应链评分。
SECURITY.md - 被动发现,第1-3个月。 处理没有经常性支出就能复利的界面:找到该细分市场中扇入最高的中心项目,并针对每个项目争取采用或一等适配器,构建“被使用”墙,并通过社区盟友而不是自我提交被策展到代码托管平台的主题界面中。
- 主动播种,持续进行。 只有 1 和 2 成立后才做:按季度日历维护 3-5 个时事通讯或播客关系,附上演示或示例仓库,赞助位与编辑覆盖明确区分。
这是第2步的效率顺序重新解读为日历,时间效应决定什么并行运行:被动发现第二,因为其回报时间跨度长;主动播种最后,因为其经常性成本。唯一一个采用者条件翻转位于第1阶段内部:信任信号在个人采用的效率上排名低,在公司采用上接近顶部,所以信任姿态要么在头四周内发布,要么根本不进入计划。这是一次性成本,解锁整个买家细分市场,所以它永远不会花在驱动流量然后停滞在采购上的情况。
四个触发器改变计划:
- 企业采用尽管文档很好但仍停滞不前。 阻碍是采购,而不是知名度——在添加任何知名度渠道之前升级信任姿态。
- 星标速度高,使用和留存持平。 公告传播了,安装没有。这是入门问题;将预算转移到文档。
- 一个中心项目驱动了大部分传递性使用。 投资于该关系和归因,而不是更多的小集成。
- 法规修复请求压力激增。 正式化披露 SLA,而不是亲自吸收负载。预期来自下游公司的要求,因为欧盟 CRA 约束制造商,而不是上游维护者:报告义务从 2026 年 9 月开始,SBOM 技术文档在 2027 年 12 月可执行。
Step 6 - Make the release stream a channel
第 6 步 - 让发布流成为一个渠道
A release broadcasts to people the maintainer cannot otherwise reach, and every recipient acts on it:
- Dependency-update bots open a pull request in every downstream repository.
- Packagers' pipelines fire off the new tag.
- Feed-harvesting newsletters pick it up.
- Watchers see it in notifications.
- Evaluators read the registry's "last published" field as a liveness proxy.
Cadence is therefore a distribution decision and the changelog a distribution asset.
The cadence rules follow from that mechanism, practitioner reasoning, not measured findings:
- Commit to a predictable maximum interval ("a fix lands within N weeks"), never a mandatory minimum.
- Batch small changes into one readable release.
- Ship security fixes out of band and fast.
An empty calendar release becomes a noise pull request downstream and trains those maintainers to auto-merge or ignore the one channel reaching existing users directly.
A long silence fails the liveness proxy even on a finished project; the cheap correction is an explicit "stable, maintained, low change rate" statement in the README. Reserve named milestones - a 1.0, a rewrite, a rename - as the legitimate reason to return to awareness channels after the launch window closes.
发布向维护者无法以其他方式触达的人广播,每个接收者都会对其采取行动:
- 依赖更新机器人会在每个下游仓库中打开一个拉取请求。
- 打包者的流水线会触发新标签。
- 订阅源收集的时事通讯会拾取它。
- 观察者会在通知中看到它。
- 评估者将注册表的“最后发布”字段读取为活跃度代理。
因此,节奏是一个分发决策,变更日志是一个分发资产。
节奏规则来自该机制,从业者推理,而不是测量发现:
- 承诺一个可预测的最大间隔(“修复在 N 周内落地”),永远不是强制性的最小间隔。
- 将小改动批量放入一个可读的发布中。
- 安全修复带外快速发布。
一个空日历的发布会变成下游的噪音拉取请求,并训练那些维护者自动合并或忽略唯一直接触达现有用户的渠道。
长时间的沉默即使在一个已完成的项目上也会使活跃度代理失败;廉价的修正是在 README 中明确声明“稳定、维护中、低变更率”。保留命名的里程碑——1.0、重写、重命名——作为发布窗口关闭后返回知名度渠道的合法理由。
Step 7 - Write the plan and validate it section by section
第 7 步 - 编写计划并逐节验证
Produce a plan the maintainer can run without you. Output shape:
markdown
undefined产出一个维护者可以在没有你的情况下运行的计划。输出形状:
markdown
undefinedDistribution plan - <project>
分发计划 - <项目>
Position today audit findings, current arrival sources, baseline numbers + date
当前位置审计发现、当前到达来源、基线数字 + 日期
Adopter individual / company / both, and what each needs
采用者 个人 / 公司 / 两者,以及各自需要什么
Foundation ring channels fixed now, with the exact gaps found
基础环渠道现在修复,以及发现的精确差距
Sequence stages 1-3 with dates, and the triggers that reorder them
序列第1-3阶段,带日期,以及重新排序的触发器
Test ring 2-3 experiments: hypothesis, signal, end date, cost
测试环 2-3 个实验:假设、信号、结束日期、成本
Core channel the one sustained channel, and why it beat the others
核心渠道 一个持续渠道,以及为什么它击败了其他渠道
Refused channel, reason, and the condition that reopens each
已拒绝 渠道、原因,以及重新打开每个的条件
Release rhythm maximum interval, milestone plan, changelog owner
发布节奏 最大间隔、里程碑计划、变更日志负责人
Review sweep cadence for decaying channels, quarterly re-decision
审查扫荡 衰减渠道的节奏,季度重新决策
Threshold check each threshold, pass or fail, with the number
阈值检查 每个阈值,通过或失败,带数字
Present each section and get agreement before writing the next. A wrong adopter call propagates into every channel choice below it, and is far cheaper to catch at the second heading than the ninth. Worked plan and weak/strong entries: [references/distribution-plan-example.md](./references/distribution-plan-example.md).
逐节呈现并在写下一节之前获得同意。错误的采用者判断会传播到下面每一个渠道选择中,在第二个标题而不是第九个标题处捕获要便宜得多。工作计划和弱/强条目:[references/distribution-plan-example.md](./references/distribution-plan-example.md)。Invocation examples
调用示例
- "Downloads flat since the Show HN. Where do people find us now?" → interview, then coverage audit; expect a half-filled registry page, not a missing channel.
- "Should we get into Homebrew and Debian?" → adopter-type question first. For a language library installed inside a virtualenv, usually no; for a standalone binary aimed at company laptops, it is a downstream-packaging plan.
- "We ship a VS Code extension. Is one marketplace enough?" → no. Dual-publish to the Microsoft marketplace and Open VSX; Cursor, Windsurf, VSCodium and Gitpod default to Open VSX.
- "Our connector is on our own GitHub. Should it be in the vendor's hub?" → yes if the hub is where their users install, but read the support tier first; see catalog family 11.
- "Will an OpenSSF Scorecard above 7 grow our downloads?" → no evidence supports that. Sell it internally as procurement enablement.
- "We have 3 hours a month. Give us the plan." → capacity is the binding constraint; propose a foundation-only mix and refuse the test ring out loud.
- “自 Show HN 以来下载量持平。人们现在在哪里找到我们?” → 访谈,然后覆盖审计;预期一个半满的注册表页面,而不是缺少的渠道。
- “我们应该进入 Homebrew 和 Debian 吗?” → 先问采用者类型。对于安装在虚拟环境中的语言库,通常不;对于面向公司笔记本电脑的独立二进制文件,这是一个下游打包计划。
- “我们发布一个 VS Code 扩展。一个市场够吗?” → 不够。双发布到 Microsoft 市场和 Open VSX;Cursor、Windsurf、VSCodium 和 Gitpod 默认使用 Open VSX。
- “我们的连接器在我们自己的 GitHub 上。它应该在供应商的中心吗?” → 如果中心是他们的用户安装的地方,是的,但先阅读支持层级;参见目录家族 11。
- “OpenSSF Scorecard 高于 7 会增加我们的下载量吗?” → 没有证据支持这一点。在内部将其推销为采购使能。
- “我们每月有 3 个小时。给我们计划。” → 容量是约束条件;提出一个仅基础的组合,并大声拒绝测试环。
Measurement and pass thresholds
测量和通过阈值
Distribution rarely offers clean attribution, so measure at two levels:
- Coverage - facts you control, auditable today.
- Arrival - behaviour observed as a trend over quarters against the step-1 baseline.
Hold the plan to these thresholds and iterate until it passes. Only the first is backed by published policy; the rest are bars this skill sets, and you must say so when reporting them:
- Sourced - every stale listing found in the audit is fixed or removed before a new channel is added, because curated lists and distro repositories drop projects that read as unmaintained by published policy, not by taste.
- Adopted bar, not an industry standard - if company adoption is in scope: supply-chain score above 7 with green artifact attestations. No public standard defines a passing score; this skill adopts this bar for enterprise review (references/published-findings.md § 4), as a procurement gate, never an adoption forecast.
- Self-set baselines - three thresholds this skill sets rather than measures:
- Every registry has 100% of its metadata fields filled (description, keyword slots, categories, license, repository, docs, homepage, funding).
- Total recurring cost is at most 60% of declared monthly capacity, leaving room for sweeps and the unplanned.
- Every channel carries an owner, an hours-per-month estimate, a named signal and a review date, with no harmful-decay channel entering without a scheduled sweep.
Report the threshold check explicitly at the end. A plan failing on capacity does not need discipline - it has one channel too many.
分发很少提供干净的归因,所以在两个层面测量:
- 覆盖 - 你控制的事实,今天可审计。
- 到达 - 作为趋势相对于第1步基线观察到的行为,跨越季度。
让计划达到这些阈值并迭代直到通过。只有第一个有已发表政策支持;其余是本技能设定的标准,你在报告时必须说明这一点:
- 有来源的 - 审计中发现的每个过时列表在新渠道添加之前都被修复或删除,因为策展列表和发行版仓库会按已发表政策(而不是品味)删除读起来像未维护的项目。
- 采用的杠,不是行业标准 - 如果公司采用在范围内:供应链评分高于 7,带有绿色工件证明。没有公开标准定义及格分数;本技能在企业审查中采用这一杠(references/published-findings.md § 4),作为采购门槛,绝不是采用预测。
- 自设基线 - 本技能设定而不是测量的三个阈值:
- 每个注册表的元数据字段 100% 填写(描述、关键词槽位、类别、许可证、仓库、文档、主页、资金)。
- 总经常性成本最多为声明月度容量的 60%,为扫荡和计划外事件留出空间。
- 每个渠道都有一个负责人、每月小时数估计、命名信号和审查日期,没有有害衰减渠道在没有计划扫荡的情况下进入。
在最后明确报告阈值检查。在容量上失败的计划不需要纪律——它只是多了一个渠道。
What the evidence supports, and what it does not
证据支持什么,不支持什么
State the confidence level on every claim - the maintainer is spending scarce hours on it. references/published-findings.md holds the citations and each study's limits.
- Supported - three of four developers consider stars before adopting or contributing, and 36.7% starred a project because they already used it (Borges and Valente, JSS 2018; 791 developers). Visible usage is a real social-proof lever - but the same survey shows promotion drives stars, so a star count measures reach as much as merit.
- Supported - only 27.5% of npm packages are depended on by any other package (Wittern et al., MSR 2016; a 2016 npm-only snapshot - trust the skew, not the digit). Entering that depended-upon core is a structural advantage.
- Not supported - supply-chain scores, signing and provenance as adoption drivers (unestablished, with maturity and funding confounding every apparent correlation), and badges as a growth lever (a first badge gives only a small, non-sustained bump - Trockman et al., ICSE 2018, npm). Weight behaviour over self-report: read dependency and download data ahead of survey preference, remembering both are gameable.
对每个声明说明置信水平——维护者正在花费稀缺时间。 references/published-findings.md 包含引文和每项研究的限制。
- 支持 - 四分之三的开发者在采用或贡献之前考虑星标,36.7% 的人因为他们_已经_使用而给项目加星(Borges 和 Valente,《JSS》2018;791 名开发者)。可见的使用是一个真正的社会证明杠杆——但同一项调查显示推广驱动星标,所以星标数量衡量的是触达和优点一样多。
- 支持 - 只有 27.5% 的 npm 包被任何其他包依赖(Wittern 等人,MSR 2016;一个 2016 年 npm 独有快照——相信偏差,而不是数字)。进入那个被依赖的核心是一个结构性优势。
- 不支持 - 供应链评分、签名和来源作为采用驱动因素(未确立,成熟度和资金混淆了每个明显的相关性),以及徽章作为增长杠杆(第一个徽章只带来一个小而不可持续的提升——Trockman 等人,ICSE 2018,npm)。优先考虑行为而不是自我报告:阅读依赖和下载数据,而不是调查偏好,记住两者都可以被操纵。
Common failure modes
常见失败模式
- Spray-and-pray listings. Twenty submissions in a weekend, all stale a year later. Three maintained surfaces outperform them and cost less.
- Confusing publishing with distribution. Shipping to a registry is table stakes; being findable there, and installable through the channel the user already uses, is the work.
- Pitching trust work as growth. Signing and SBOMs remove a procurement blocker. Promising a download lift sets up a failure the next security ask will pay for.
- Reading stars as the outcome. Stars measure attention and promotion reach; downloads, dependents, cloners and issues from strangers measure distribution.
- Planning past the capacity. Capacity shrinks when the project succeeds - support load grows first. Plan for the worst realistic month.
- Never retiring anything. A quarterly review with the power to kill a channel keeps the portfolio inside capacity.
Per-family traps - self-submitting to curated code-host surfaces, treating a downstream package as owned, cadence theater, one description copy-pasted everywhere, trusting external trending tools - are listed against the family they belong to in references/channel-catalog.md.
- 喷洒并祈祷的列表。 一个周末二十个提交,一年后都过时了。三个维护良好的界面胜过它们,成本更低。
- 混淆发布与分发。 发货到注册表是入场券;在那里可被发现,并通过用户已经使用的渠道可安装,才是工作。
- 把信任工作推销为增长。 签名和 SBOM 消除采购障碍。承诺下载提升会为下一次安全问题的代价埋下失败。
- 把星标读作结果。 星标衡量注意力和推广触达;下载、依赖者、克隆者和陌生人的 issues 衡量分发。
- 超出容量规划。 当项目成功时容量会缩小——支持负载首先增长。为最坏的实际月份做计划。
- 从不退役任何东西。 一个有权终止渠道的季度审查使组合保持在容量内。
每个家族的陷阱——自我提交到策展的代码托管界面、将下游软件包视为己有、节奏表演、一个描述到处复制粘贴、信任外部趋势工具——在 references/channel-catalog.md 中列在它们所属的家族下。
References
参考
- references/channel-catalog.md - the eleven families: what each rewards, its cost, decay profile, traps.
- references/coverage-audit.md - inventory, metadata checks, coverage and arrival metrics.
- references/distribution-plan-example.md - worked plan, weak vs strong entries, refusals.
- references/published-findings.md - the studies and policies behind every sourced claim, and what each does not prove.
- - the repository surface every channel points at.
samber/developer-relations-skills@readme-optimization - - the release notes this channel broadcasts.
samber/developer-relations-skills@changelog-writing - - execution for family 10 (machine and agent-readable surfaces).
samber/developer-relations-skills@coding-agent-docs-optimization
- references/channel-catalog.md - 十一个家族:每个奖励什么、成本、衰减概况、陷阱。
- references/coverage-audit.md - 清单、元数据检查、覆盖和到达指标。
- references/distribution-plan-example.md - 工作计划、弱与强条目、拒绝。
- references/published-findings.md - 每个有来源声明背后的研究和政策,以及每个不证明什么。
- - 每个渠道指向的仓库界面。
samber/developer-relations-skills@readme-optimization - - 此渠道广播的发布说明。
samber/developer-relations-skills@changelog-writing - - 家族 10(机器和代理可读界面)的执行。
samber/developer-relations-skills@coding-agent-docs-optimization