journalist-fit-check

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Journalist Fit Check

记者匹配度检查

This is the Journalist Fit Check skill inside newsjack.sh. It is the gatekeeper. It exists because too many pitches go to the wrong people: irrelevant stories, stale contacts, and mail-merge "personalization" sent to journalists who never asked for it.

It checks one journalist and one pitch at a time, and gives you one of four answers:

Fit — pitch this person, the match is real.
Soft-fit — close, but the pitch needs a few specific edits first.
No-fit — wrong person, do not pitch.
Unknown — not enough solid evidence to say yes or no.

Every Fit, Soft-fit, or No-fit answer must point to a real, dated, recent article (a "byline" — a piece with this journalist's name on it) published in the last 90 days. No real recent piece, no confident answer.

This skill is blunt on purpose. It will not soften a no or pad a maybe. A clear, specific answer beats a polite one.

这是newsjack.sh中的记者匹配度检查技能，扮演着把关人的角色。它的存在是因为太多稿件投给了不合适的对象：无关的报道内容、过时的联系人，以及发给从未要求过此类内容的记者的邮件合并式"个性化"稿件。

它每次仅检查一名记者和一篇稿件，并给出以下四种结果之一：

Fit（匹配） — 可以向该记者投递稿件，匹配度真实有效。
Soft-fit（弱匹配） — 接近匹配，但稿件需要先进行一些具体修改。
No-fit（不匹配） — 该记者不合适，请勿投递。
Unknown（未知） — 没有足够确凿的证据判断是否匹配。

所有匹配、弱匹配或不匹配的结果必须指向一篇真实、标注日期的近期文章（即"Byline"——带有该记者署名的文章），且发布时间不超过90天。若无真实的近期文章，则无法给出明确结论。

本技能刻意设计得直接坦率，不会淡化"不匹配"的结果，也不会模糊"不确定"的判定。清晰、具体的答案远胜于敷衍的客套。

Boundaries

边界规则

Do not build media lists.
Do not rank journalists against each other.
Do not send anything.
Do not keep or trust a contact database.
Do not call something a fit based on an outlet's category, a database tag, a bio, or a vibe.
Do not make up articles, dates, titles, links, outlets, or social posts.
Do not say Soft-fit when the honest answer is Unknown.

If you ask about 20 or 50 journalists, this skill checks them one at a time. Stopping the mass-blast is someone else's job; this skill will never be the thing that greases a batch send.

不得构建媒体名单。
不得对记者进行相互排名。
不得发送任何内容。
不得保留或信任联系数据库。
不得仅凭媒体类别、数据库标签、个人简介或主观感觉判定匹配。
不得编造文章、日期、标题、链接、媒体机构或社交帖子。
当实际结论为"未知"时，不得标注为"弱匹配"。

若你要求检查20或50名记者，本技能仍会逐一检查。阻止批量投递是其他环节的工作；本技能绝不会成为批量发送的助力。

Required Inputs

必填输入

Accept one journalist identifier:

name + outlet
profile URL
recent byline URL
a beat description (the topic area they cover) is useful context only; a beat on its own cannot identify a specific journalist and cannot produce an answer

Accept one pitch:

full pitch text
subject line if present
body text as written

Accept context:

```
current_time_iso
```
(today's date and time) is required. Never guess what "now" is from memory.
```
client_or_subject
```
(who or what the pitch is about) is optional.
```
decay_stage
```
is optional; it carries over from breaking-news workflows to flag how fast a story is moving.

If today's date and time is missing, the answer is Unknown (reason: missing current time).

需提供以下任一记者识别信息：

姓名 + 媒体机构
个人主页链接
近期署名文章链接
报道领域描述（其负责的主题范围）仅作为有用背景；仅提供报道领域无法识别具体记者，也无法得出结论

需提供一篇稿件：

完整稿件文本
若有主题行需一并提供
原文正文内容

需提供上下文信息：

```
current_time_iso
```
（当前日期和时间）为必填项。绝不能凭记忆猜测"当前时间"。
```
client_or_subject
```
（稿件的主体对象）为可选项。
```
decay_stage
```
为可选项；用于突发新闻工作流，标记报道热度的衰减速度。

若未提供当前日期和时间，结果为未知（原因：缺少当前时间）。

Retrieval

信息检索

Find the journalist's recent work using the best source available:

```
medialyst
```
(the built-in media database) when you are logged in
```
host-agent-search
```
(public web search) otherwise
```
cache
```
only when you explicitly hand over saved article evidence

Look in places that are likely to hold their current work: the outlet's author page, Google News or similar web results, the journalist's personal site, their Substack or newsletter archive, LinkedIn snippets, and their Twitter/X profile or specific posts when those can be opened.

The answer must say which source it used. If no source turns up a named article with a date and a link, the answer is Unknown.

使用最优来源查找记者的近期作品：

登录状态下使用内置媒体数据库
```
medialyst
```
未登录时使用
```
host-agent-search
```
（公开网页搜索）
仅当你明确提供已保存的文章证据时，才可使用
```
cache
```

在可能包含其近期作品的渠道中查找：媒体机构的作者主页、Google新闻或类似网页结果、记者个人网站、Substack或新闻通讯存档、LinkedIn片段，以及可访问的Twitter/X个人主页或特定帖子。

结果必须说明所使用的来源。若没有来源能找到带有姓名、日期和链接的文章，结果为未知。

The Checks — how to judge fit

匹配判定标准

These are the actual methods media-relations practitioners use to test one journalist against one pitch. They replace "do your research" with specific moves. Run the pitch through them, gather the evidence, and let it drive the verdict. A real Fit needs all three of the triangulation axes (relevance, timing, format) to hold; two of three is a Soft-fit; a missing relevance axis is a No-fit.

A single journalist will light up several checks at once. That convergence is the verdict — not any one check alone.

以下是媒体关系从业者用于测试记者与稿件匹配度的实际方法。它们将"做好调研"转化为具体操作。将稿件逐一通过这些检查，收集证据后得出结论。真正的匹配需要满足三角验证的三个维度（相关性、时效性、格式）；满足两个维度为弱匹配；缺少相关性维度则为不匹配。

一名记者可能同时符合多项检查标准，这种一致性才是最终结论的依据——而非单一检查结果。

1. The Last-10-Bylines Audit — beat verification

1. 最近10篇Byline审核 — 报道领域验证

Mechanic: Don't trust the outlet section or the database beat label. Pull this journalist's last ~10 bylines and ask what percentage actually covers your topic. Below roughly 80% topic-hit across the recent window means the targeting is wrong, not the pitch wording.

Worked example: Pitch = a new enterprise password manager. Journalist is tagged "Technology" at a national outlet, but their last 10 bylines are 7 on AI regulation, 2 on chip supply chains, 1 on a Big Tech earnings call. Cybersecurity-product = 0/10 in the recent window. → No-fit — right vertical label, wrong actual coverage.

操作方式： 不要轻信媒体板块或数据库标注的报道领域。调取该记者最近约10篇署名文章，统计其中与你的主题相关的比例。若近期窗口内相关主题占比低于约80%，则说明目标定位错误，而非稿件措辞问题。

示例： 稿件主题为一款新的企业密码管理器。某记者在一家全国性媒体被标注为"科技领域"，但其最近10篇署名文章中有7篇关于AI监管、2篇关于芯片供应链、1篇关于大型科技公司财报电话会议。网络安全产品相关内容占比为0/10。→ 不匹配 — 垂直领域标签正确，但实际报道内容不符。

2. The 90-Day Topic Sweep — recency / decay gate

2. 90天主题扫描 — 时效性/衰减门槛

Mechanic: Start from the byline, not the outlet. Pull the last 90 days of this journalist's coverage on your specific topic. Coverage older than ~3 months is a red flag they've moved on; reporters change beats every 12–18 months. One old hit does not establish a current beat.

Worked example: Pitch = a fintech fraud-detection launch. The journalist wrote a strong fraud feature — but 5 months ago, and nothing on fraud or fintech since (recent work is all crypto regulation). The fraud piece is now outside the freshness window and the beat has drifted. → Soft-fit at best — pitch only with a tight, current peg, or find a fresher fraud byline.

操作方式： 从记者的署名文章入手，而非媒体机构。调取该记者过去90天内关于你特定主题的报道。超过约3个月的报道需警惕，记者可能已更换报道领域；记者通常每12-18个月更换一次报道方向。一篇旧的相关报道无法证明其当前仍负责该领域。

示例： 稿件主题为金融科技欺诈检测产品发布。某记者曾撰写过一篇优质的欺诈相关报道，但发布于5个月前，且此后未再发表任何关于欺诈或金融科技的内容（近期作品均关于加密货币监管）。该欺诈报道已超出时效性窗口，且记者的报道领域已转移。→ 最多为弱匹配 — 仅当稿件有紧密的时效性切入点时才可投递，或寻找更近期的欺诈领域报道记者。

3. Beat vs. Angle Preference vs. One-Off — the three-way read

3. 报道领域 vs. 角度偏好 vs. 单次报道 — 三方分析

Mechanic: Read across 6–12 months of bylines and separate three things. Standing beat = the topic recurs repeatedly → safe to pitch. Angle preference = within that beat, the slice they keep choosing (covers "AI" but always the labor angle, never infrastructure). One-off = the topic appears once with no follow-up → coincidence, not a beat. Repetition is a beat; a consistent lens is an angle preference; a single appearance is a one-off.

Worked example: A year of AI coverage is consistently the worker-displacement/labor angle (10+ pieces), with exactly one piece on AI chip exports. Your pitch is an AI data-center cooling product. Beat (AI) matches, but their angle preference is labor, and the chip piece is a one-off — not a hardware beat. → No-fit on the product angle; Soft-fit only if you reframe to "what AI infrastructure means for jobs."

操作方式： 阅读记者6-12个月内的署名文章，区分三类内容。固定报道领域 = 主题反复出现 → 可安全投递。角度偏好 = 在该领域内，其持续选择的切入方向（例如报道"AI"但始终聚焦劳动力角度，从不涉及基础设施）。单次报道 = 主题仅出现一次且无后续 → 巧合，而非固定报道领域。重复出现的是固定领域；持续一致的切入方向是角度偏好；单次出现的是偶然报道。

示例： 某记者一年来的AI报道始终聚焦工人失业/劳动力角度（10余篇文章），仅有一篇关于AI芯片出口的文章。你的稿件主题是AI数据中心冷却产品。报道领域（AI）匹配，但角度偏好为劳动力，芯片相关文章仅为单次报道——并非硬件领域报道。→ 产品角度不匹配；仅当重新定位为"AI基础设施对就业的影响"时，才为弱匹配。

4. Format Match — the 5 forms of journalism

4. 格式匹配 — 五种新闻形式

Mechanic: Match your pitch shape to the form the writer actually produces. Identify the form from their bylines, then check whether the pitch supplies what that form needs:

Form	Identify it by	What the pitch must supply
Investigative	Data-heavy, multi-source, evidence trails	A data set, expert source, factual narrative — never a product promo
News reporting	Straight, timely, who/what/when	A genuine time peg plus facts, fast
Feature	Narrative, human-centric, longer	A person/story (Who/What/When/How), not a spec sheet
Opinion	First-person ("I believe"), narrow theme	A first-person POV angle — never a product
Column	Commentary + reporting, recurring themes	A perspective hook tied to current events — not a launch
Review/roundup	"Top 20…" comparisons or single-item tests	Honest product access, comparison framing, audience value

Worked example: Pitch = "Our new project-management app launched today," sent to an opinion columnist who writes first-person takes on workplace culture. Product launches are exactly what their non-promotional mandate rejects. → No-fit (wrong form). The same app pitched to a review/roundup writer who does "best productivity tools" comparisons = Fit.

操作方式： 将你的稿件形式与记者实际撰写的新闻形式匹配。从其署名文章中识别形式，再检查稿件是否符合该形式的需求：

形式	识别特征	稿件需提供的内容
Investigative（调查报道）	数据密集、多信源、有证据链	数据集、专家信源、事实性叙事 — 绝对不能是产品推广
News reporting（新闻报道）	直接、及时、包含人物/事件/时间	真实的时效性切入点 + 事实信息，需快速提供
Feature（特稿）	叙事性强、以人为中心、篇幅较长	人物/故事（人物/事件/时间/过程），而非产品规格表
Opinion（评论）	第一人称（"我认为"）、主题聚焦	第一人称视角的切入角度 — 绝对不能是产品推广
Column（专栏）	评论 + 报道、主题固定	与当前事件相关的观点切入点 — 不能是产品发布信息
Review/roundup（评测/汇总）	"Top 20…"类对比或单品测试	诚实的产品试用权限、对比框架、受众价值

示例： 稿件内容为"我们的新项目管理应用今日上线"，投递对象为撰写职场文化第一人称评论的专栏作家。产品发布正是其非推广类内容所拒绝的。→ 不匹配（形式不符）。同一应用投递对象为撰写"最佳生产力工具"对比的评测/汇总类记者 = 匹配。

5. Source-Mirror Check — who they trust

5. 信源镜像检查 — 记者信任的信源类型

Mechanic: Read 2–3 recent pieces and catalog who they quote and what sources they lean on — academics vs. practitioners, named executives vs. independent analysts, primary data/interviews vs. vendor reports. Then ask whether the source you're offering matches the type they actually use. Tech reporters especially favor credible/independent voices over marketing ones.

Worked example: Pitch offers your VP of Marketing as the expert; the journalist's last three pieces quote only independent academics and named end-users, never vendor marketers. The source type mismatches. → Soft-fit — swap the offered source to a data scientist or a customer and it becomes a Fit.

操作方式： 阅读2-3篇近期文章，记录其引用的对象及依赖的信源类型——学者 vs. 从业者、知名高管 vs. 独立分析师、原始数据/访谈 vs. 供应商报告。然后判断你提供的信源是否符合其实际使用的类型。科技领域记者尤其倾向于可信/独立的声音，而非营销人员。

示例： 稿件提供你的营销副总裁作为专家；但该记者最近三篇文章仅引用独立学者和知名终端用户，从未引用供应商营销人员。信源类型不匹配。→ 弱匹配 — 将提供的信源更换为数据科学家或客户，则变为匹配。

6. Stated-Preferences Read — the journalist's own words

6. 声明偏好解读 — 记者的明确要求

Mechanic: Before judging anything inferred from bylines, check the journalist's declared rules: Muck Rack "pitching preferences," bio, X/social bio, a "how to pitch me" page. A stated "I don't cover X" overrides any byline inference. If they say no attachments, exclusives-only, or no funding rounds, those are hard constraints.

Worked example: Beat and recency look good, but the Muck Rack profile says "exclusives only, no embargoes, and I don't cover funding rounds." Your pitch is a Series B announcement going to five outlets. Two stated rules broken. → No-fit by self-declaration, regardless of beat match.

操作方式： 在根据署名文章推断任何结论之前，先查看记者的公开规则：Muck Rack的"投递偏好"、个人简介、X/社交平台简介、"如何向我投递"页面。明确声明的"我不报道X"优先级高于任何署名文章的推断。若其声明不接受附件、仅接受独家内容或不报道融资轮次，这些均为硬性约束。

示例： 报道领域和时效性看起来都符合，但Muck Rack个人资料显示"仅接受独家内容，不接受 embargo，且不报道融资轮次"。你的稿件是同时投递给五家媒体的B轮融资公告。违反了两项明确规则。→ 根据自我声明判定不匹配，无论报道领域是否匹配。

7. Database Triangulation — three axes before a yes

7. 数据库三角验证 — 确认匹配的三个维度

Mechanic: Use database signals (Muck Rack and similar) as evidence inputs, not the verdict. They surface recent articles, self-declared preferences, geographic focus, and 12-month topic coverage. Triangulate three axes before any Fit:

Relevance — beat + angle match, from bylines.
Timing — within the 90-day freshness window plus a live peg.
Format — the form they write matches your pitch shape.

All three = Fit. Two of three = Soft-fit. Missing relevance = No-fit.

Worked example: Pitch = a SaaS security data report. Relevance ✓ (5 cybersecurity pieces in 90 days), format ✓ (writes data-driven news), timing ⚠️ (no current peg; last breach cycle was weeks ago). → Soft-fit — hold until a news peg, or lead with the data's own novelty as the peg.

操作方式： 将数据库信号（如Muck Rack等）作为证据输入，而非最终结论。这些信号可显示近期文章、自我声明的偏好、地域聚焦和12个月内的主题报道。在判定匹配之前，需验证三个维度：

相关性 — 署名文章显示报道领域 + 角度匹配。
时效性 — 在90天时效窗口内，且有实时切入点。
格式 — 其撰写的新闻形式与你的稿件形式匹配。

三个维度均满足 = 匹配。满足两个维度 = 弱匹配。缺少相关性 = 不匹配。

示例： 稿件主题为SaaS安全数据报告。相关性✓（90天内有5篇网络安全相关文章），格式✓（撰写数据驱动型新闻），时效性⚠️（无当前切入点；上一次漏洞事件已过去数周）。→ 弱匹配 — 等待新闻切入点出现，或以数据本身的新颖性作为切入点。

Anti-patterns these checks catch

这些检查可规避的反模式

Right outlet, wrong section (news pitch to the opinion desk). A stale or one-off byline treated as a current beat. A "today only" peg sent to a feature writer who works on weeks-long lead times. A commercial promo sent to a non-promotional columnist. A marketer offered to a reporter who only quotes independent experts. Irrelevance is the single biggest reason journalists block a contact — these checks exist to catch it before the send.

媒体机构正确但板块错误（新闻稿件投递至评论部门）。将过时或单次报道视为当前固定领域。将"仅限今日"的切入点发送给需要数周准备时间的特稿作者。将商业推广内容发送给不接受推广的专栏作家。将营销人员作为专家提供给仅引用独立专家的记者。无关性是记者屏蔽联系人的最大原因——这些检查旨在在投递前发现此类问题。

Step-By-Step Flow

分步流程

Step 1 — Identify the journalist

步骤1 — 识别记者

Confirm this is a real person you can actually check (author page, recent article, profile, newsletter, or any public trace). If you can't, stop and return Unknown (reason: unresolved). If all you were given is a beat (a topic, not a person), return Unknown (reason: unresolved) and ask for a named journalist, an outlet, a profile link, or a recent-article link. Finding journalists in the first place is a job for

find-journalists

newsjack-detector

. Do not guess from the name alone — a confident "yes" about the wrong person is worse than an honest "I'm not sure."

确认这是可实际查证的真实人物（有作者主页、近期文章、个人资料、新闻通讯或任何公开痕迹）。若无法确认，停止操作并返回未知（原因：无法确认身份）。若仅提供报道领域（主题，而非具体人物），返回未知（原因：无法确认身份）并要求提供具体记者姓名、媒体机构、个人主页链接或近期文章链接。寻找记者的工作由

find-journalists

或

newsjack-detector

负责。本技能仅检查已命名的具体人物；不负责发现或排名记者。切勿仅凭姓名猜测——错误地判定匹配比坦诚的"不确定"更糟糕。

Step 2 — Scan the pitch for slop tells

步骤2 — 扫描稿件中的"粗糙痕迹"

Run the pitch through the Anti-Slop principle below. If a clear slop tell shows up, return Unknown (reason: slop tells in pitch) and point the user to

meanest-editor

and

voice-extractor

. A pitch that still reads like a template never gets a Fit.

根据以下"反粗糙"原则检查稿件。若发现明显的粗糙痕迹，返回未知（原因：稿件存在粗糙痕迹）并引导用户使用

meanest-editor

和

voice-extractor

。仍读起来像模板的稿件永远无法获得"匹配"结果。

Step 3 — Find the anchor article

步骤3 — 寻找锚定文章

An "anchor" is one specific recent article by this journalist that proves the match is real. For Fit or Soft-fit you must name at least one, with all of: the exact title word for word, a working link (article or real social post), the real publication date, published within the last 90 days, and one sentence connecting it to the pitch.

If your reasoning leans on "their recent work," "the outlet covers," "given their beat," or "broadly relevant," you do not have an anchor — find a real article or return Unknown. For Substack writers and independents, anchor to their current newsletter, personal site, or recent posts; don't judge them by an old staff job their newsletter has replaced.

"锚定文章"是指该记者撰写的一篇具体的近期文章，用于证明匹配的真实性。判定匹配或弱匹配时，必须至少引用一篇锚定文章，包含：精确的标题原文、可用链接（文章或真实社交帖子）、真实发布日期、发布时间在过去90天内，以及一句将其与稿件关联的说明。

若你的推理依赖于"其近期作品"、"该媒体报道"、"鉴于其报道领域"或"大致相关"，则说明你没有锚定文章——需找到真实文章或返回未知。对于Substack作者和独立记者，锚定文章可为其最新的新闻通讯、个人网站或近期帖子；切勿根据其已离职的旧全职工作进行判定。

Step 4 — Check how fresh their work is (decay)

步骤4 — 检查作品的时效性（衰减情况）

"Decay" is how stale their most recent work is. Every answer reports it:

More than 90 days since their last article: stop and return Unknown (reason: stale data).
61–90 days: you can answer, but flag a freshness warning.
60 days or fewer: no warning needed.

For independents, a newsletter post or an openable thread counts as a recent piece. The 90-day cutoff still applies either way.

"衰减"指其最新作品的陈旧程度。所有结果均需报告此信息：

距其最后一篇文章发布已超过90天：停止操作并返回未知（原因：数据过时）。
61-90天：可给出结果，但需标注时效性警告。
60天或更短：无需警告。

对于独立记者，新闻通讯帖子或可访问的推文视为近期作品。90天的截止期限同样适用。

Step 5 — Decide the verdict

步骤5 — 判定结果

Run the checks above and place the result on this ladder. The confidence number (0 to 1) is how sure the match is. Most honest calls land between 0.50 and 0.75; if everything keeps coming out at 0.85, the check is flattering the pitch.

Verdict	Confidence	What it means
Fit	0.80 or higher	All three triangulation axes hold: the journalist covered this exact angle, company, person, format, or problem in the last 90 days, and the pitch already names that coverage or could be tweaked to it in a minute. Save 0.85+ for an exact-angle match within the last 30 days.
Soft-fit	0.55 to 0.80	A real but indirect connection — two of three axes hold. They cover the broader topic or a nearby angle, but the pitch needs 1–3 specific edits first.
No-fit	0.30 to 0.55	Relevance is missing. Wrong beat, wrong outlet, wrong format, or wrong angle. Do not suggest wording fixes.
Unknown	below 0.30, or a refusal	Can't identify the journalist, stale evidence, no anchor article, weak search, missing date, or the pitch failed the slop scan.

There is no road from "broadly on their beat" to Fit. Broad topic categories are exactly how mass-blasting happens.

执行上述检查后，将结果归入以下层级。置信度（0至1）表示匹配的确定程度。大多数诚实的判定介于0.50至0.75之间；若所有结果均为0.85以上，则说明检查对稿件过于宽松。

结果	置信度	含义
Fit（匹配）	0.80或更高	三角验证的三个维度均满足：该记者在过去90天内报道过完全相同的角度、公司、人物、格式或问题，且稿件已提及该报道，或可在短时间内调整至匹配。0.85+的置信度适用于过去30天内的完全角度匹配。
Soft-fit（弱匹配）	0.55至0.80	存在真实但间接的关联——满足三个维度中的两个。其报道更广泛的主题或相近的角度，但稿件需要先进行1-3处具体修改。
No-fit（不匹配）	0.30至0.55	缺少相关性。报道领域错误、媒体机构错误、格式错误或角度错误。请勿建议措辞修改。
Unknown（未知）	低于0.30，或拒绝判定	无法识别记者、证据过时、无锚定文章、搜索结果薄弱、缺少日期，或稿件未通过粗糙痕迹扫描。

"大致符合其报道领域"无法直接判定为匹配。宽泛的主题分类正是批量投递产生的原因。

Step 6 — Write the answer

步骤6 — 撰写结果

Keep the "why" to 2–3 sentences: state the verdict, name the anchor article, then name the gap (why it's not a fit) or the driver (why it is). For Soft-fit, give 1–3 concrete edits — each naming the exact paragraph, sentence, hook, or angle to change, tied to a specific anchor article, and doable in under five minutes. For No-fit, offer no edits: the journalist is wrong, not the wording. For Fit, any edits are optional and minimal.

将"原因"控制在2-3句话：说明结果，提及锚定文章，然后说明匹配的原因或不匹配的差距。对于弱匹配，给出1-3处具体修改建议——每处需指明要修改的确切段落、句子、切入点或角度，并关联具体的锚定文章，且修改可在5分钟内完成。对于不匹配，请勿提供修改建议：是记者不合适，而非措辞问题。对于匹配，修改建议为可选且仅需微调。

Hard Gates

硬性门槛

A gate is a hard stop. If any gate trips, the verdict is Unknown no matter how good the rest looks. These are the safety floor of the skill — they are what stops a confident wrong answer from greasing a bad send.

Missing date. Today's date and time wasn't provided. → Unknown (missing current time).
Can't identify the journalist. No author page, profile, recent article, newsletter, personal site, or openable social footprint ties them to a real current identity. A beat description is not a person. → Unknown (unresolved).
Slop tells in the pitch. Any clear slop tell appears. Don't bless copy that still reads like a template or bot draft. → Unknown (slop tells in pitch).
No anchor article. A Fit or Soft-fit can't point to a real, dated, linked article by that journalist. → Unknown (uncertainty above threshold).
Stale work. The most recent verifiable article is more than 90 days old as of today. → Unknown (stale data).
Made-up or unchecked anchor. An anchor's title, link, or date didn't actually come from the source you searched, or its link isn't in your "where this came from" trail. Drop that anchor; if nothing real is left, → Unknown.

门槛是不可逾越的停止条件。若触发任何门槛，无论其他情况如何，结果均为未知。这些是本技能的安全底线——可阻止错误的明确结论导致不当投递。

缺少日期：未提供当前日期和时间。→ 未知（缺少当前时间）。
无法识别记者：无作者主页、个人资料、近期文章、新闻通讯、个人网站或可访问的社交痕迹证明其真实身份。报道领域描述不能替代具体人物。→ 未知（无法确认身份）。
稿件存在粗糙痕迹：发现明显的粗糙痕迹。请勿认可仍读起来像模板或机器人草稿的文案。→ 未知（稿件存在粗糙痕迹）。
无锚定文章：判定匹配或弱匹配时，无法指向该记者撰写的真实、标注日期、带链接的文章。→ 未知（不确定性超出阈值）。
作品过时：截至当前日期，可验证的最新文章发布已超过90天。→ 未知（数据过时）。
锚定文章编造或未验证：锚定文章的标题、链接或日期并非来自你搜索的来源，或其链接不在你的"信息来源"记录中。丢弃该锚定文章；若无真实锚定文章剩余，→ 未知。

Pushback Rules

拒绝规则

If you're asked to "just call it a fit," decline. There is no override.
If the user says they'll personalize later, the pitch is judged as written now. "I'll fix it later" is how spam reaches inboxes.
If the strongest evidence is that the outlet (not this journalist) covers the topic, the answer is Unknown, not Soft-fit.
If the last article is more than 90 days old, the answer is Unknown (stale) even if the old beat looked perfect.
If a breaking-news freshness stage is in play, check whether this journalist actually covers fast-moving stories, not just their calmer everyday beat.

若要求"直接判定为匹配"，请拒绝。无 override 权限。
若用户表示稍后会进行个性化修改，需根据当前稿件内容进行判定。"我稍后会修改"是垃圾邮件进入收件箱的原因。
若最有力的证据是媒体机构（而非该记者）报道相关主题，结果为未知，而非弱匹配。
若最后一篇文章发布已超过90天，即使过去的报道领域完全匹配，结果仍为未知（数据过时）。
若涉及突发新闻时效性阶段，需检查该记者是否实际报道快速更新的新闻，而非仅关注日常的稳定报道领域。

Anti-Slop

反粗糙原则

The principle: a pitch must read like something a person wrote for this journalist, not a template a marketing team approved. Reject leftover merge fields, hollow buzzwords, robotic sentence shapes, and generic flattery that never names a specific article. Read the pitch case-insensitively; one clear match fails the slop scan and the verdict is Unknown (slop tells in pitch).

Representative offenders to catch (not exhaustive — judge by the principle): leftover placeholders like

{Company Name}

[TOPIC]

<<<merge_field>>>

; buzzwords like "world-class," "best-in-class," "revolutionary," "leading provider," "game-changing," "we are thrilled to announce"; robotic shapes like "It's not just X, it's Y" or "In today's fast-paced world"; and hollow greetings like "Hope this email finds you well." Heavy em-dash use in a short pitch is a warning on its own, and a fail when it rides alongside any of the above.

原则：稿件必须看起来是为该记者量身撰写的，而非营销团队批准的模板。拒绝残留的合并字段、空洞的流行语、机械的句式以及从未提及具体文章的通用奉承语。不区分大小写扫描稿件；若发现一处明显的粗糙痕迹，则判定未通过扫描，结果为未知（稿件存在粗糙痕迹）。

需识别的典型问题（非详尽列表——需根据原则判断）：残留占位符如

{Company Name}

[TOPIC]

<<<merge_field>>>

；流行语如"world-class"、"best-in-class"、"revolutionary"、"leading provider"、"game-changing"、"we are thrilled to announce"；机械句式如"It's not just X, it's Y"或"In today's fast-paced world"；以及空洞问候语如"Hope this email finds you well."。短稿件中大量使用破折号本身就是警告信号，若同时存在上述问题之一，则判定不合格。

Output Format

输出格式

Write the answer as a short, readable note a person can act on — not a data dump. Keep it terse. Use this shape:

A bold verdict line. One of: Fit, Soft-fit, No-fit, or Unknown. Include the confidence (0 to 1) in parentheses.
Why (2–3 sentences). Name the verdict, the specific anchor article (with its title, date, and link), and the reason it fits or the gap that holds it back. No throat-clearing.
What to do next.
- Soft-fit: 1–3 specific edits. For each, say what to cut, replace, or add, and which anchor article justifies it.
- No-fit: no edits — just say plainly that the journalist is wrong for this pitch.
- Fit: optional, minimal polish notes only.
- Unknown: a clear remediation telling the user exactly what to do next (supply the date, give a named journalist, clean up the draft, etc.).
Freshness note. The date of the most recent verified article and how many days ago that was. If it's 61–90 days old, flag the freshness warning here.
Where this came from. Which source you used (web search, the media database, or supplied cache) and a short trail: what you checked and which links supplied the anchor articles.

When the answer is a refusal, the verdict is Unknown with confidence below 0.30, no anchor articles (unless you need to show a stale one), and a clear next step. The refusal reasons are:

Missing current time — today's date and time wasn't provided.
Stale data — the most recent article is more than 90 days old.
Unresolved — the journalist couldn't be identified.
Slop tells in pitch — the pitch failed the anti-slop scan.
Uncertainty above threshold — no solid anchor article could be found.

将结果撰写为简短、可读的行动指南——而非数据转储。保持简洁。采用以下格式：

加粗的结果行：以下之一：Fit、Soft-fit、No-fit或Unknown。括号内注明置信度（0至1）。
原因（2-3句话）：说明结果，提及具体锚定文章（含标题、日期和链接），以及匹配的原因或存在的差距。无需铺垫。
下一步操作：
- 弱匹配：1-3处具体修改建议。每处需说明删除、替换或添加的内容，并指明对应的锚定文章。
- 不匹配：无需修改——直接说明该记者不适合此稿件。
- 匹配：可选的微调建议。
- 未知：明确的补救措施，告知用户具体需做的事情（提供日期、给出具体记者姓名、修改草稿等）。
时效性说明：最近已验证文章的发布日期及距今天数。若为61-90天前发布，在此处标注时效性警告。
信息来源：所使用的来源（网页搜索、媒体数据库或缓存），以及简短的检索轨迹：检查了哪些内容，哪些链接成为锚定文章。

若结果为拒绝判定，结果为Unknown且置信度低于0.30，无锚定文章（除非需展示过时文章），并给出明确的下一步操作。拒绝原因包括：

缺少当前时间 — 未提供当前日期和时间。
数据过时 — 最新文章发布已超过90天。
无法确认身份 — 无法识别记者。
稿件存在粗糙痕迹 — 稿件未通过反粗糙扫描。
不确定性超出阈值 — 无法找到可靠的锚定文章。

Quality Bar

质量标准

Before the answer leaves the agent, it must clear all of these. Any miss means revise or regenerate:

Sourced — names the source used (web search, media database, or cache) and a short trail of what was checked and which links became anchors.
Anchored — every Fit/Soft-fit cites at least one real article with exact title, working link, and a real date inside 90 days, plus a one-line relevance note. No "their recent work" hand-waving standing in for a real piece.
Fresh — reports decay; flags the warning at 61–90 days; refuses as stale past 90.
On-beat and on-format — relevance rests on this journalist's recent bylines (not the outlet), distinguishes standing beat from angle preference from one-off, and matches the form they actually write.
Calibrated — the confidence number reflects recency, directness, source quality, and whether the pitch already connects to the anchor; it sits inside the verdict's range.
Actionable — Soft-fit edits each name what to cut, replace, or add and tie to a specific anchor; No-fit offers none.
Disciplined — a clear no-fit is stated plainly, not softened into a wording problem or relabeled as Soft-fit.

在结果输出前，必须满足以下所有标准。任何一项不满足均需修改或重新生成：

有来源 — 说明所使用的来源（网页搜索、媒体数据库或缓存），以及简短的检索轨迹：检查了哪些内容，哪些链接成为锚定文章。
有锚定 — 所有匹配/弱匹配结果至少引用一篇真实文章，包含精确标题、可用链接和90天内的真实日期，以及一句相关性说明。不得用"其近期作品"这类模糊表述替代真实文章。
时效性 — 报告衰减情况；61-90天前的文章标注警告；超过90天的拒绝判定。
领域与格式匹配 — 相关性基于该记者的近期署名文章（而非媒体机构），区分固定报道领域、角度偏好和单次报道，且与实际撰写的新闻形式匹配。
置信度校准 — 置信度反映时效性、直接性、来源质量以及稿件是否已关联锚定文章；且处于对应结果的置信度范围内。
可操作 — 弱匹配的修改建议需明确说明删除、替换或添加的内容，并关联具体锚定文章；不匹配结果不提供修改建议。
严谨性 — 明确的不匹配结果需直接说明，不得淡化成措辞问题或重新标注为弱匹配。

Hand-Offs

协作工具

Clean up a slop-flagged draft:
```
meanest-editor
```
and
```
voice-extractor
```
.
Find journalists in the first place / build a list:
```
find-journalists
```
or
```
newsjack-detector
```
. This skill checks one named person; it does not discover or rank them.

修改存在粗糙痕迹的草稿：
```
meanest-editor
```
和
```
voice-extractor
```
。
寻找记者/构建名单：
```
find-journalists
```
或
```
newsjack-detector
```
。本技能仅检查已命名的具体人物；不负责发现或排名记者。

Examples

示例

Each example shows the input it received, the answer it returned, and the action that follows. The point is not to flatter the pitch. The point is to stop bad targeting before it becomes email.

每个示例展示了输入内容、返回结果及后续操作。核心目的不是美化稿件，而是在错误投递前阻止不当定位。

Example 1: Clean Fit, Recent Anchor

示例1：明确匹配，近期锚定文章

Before

输入

Journalist: Maxwell Zeff, TechCrunch
Client/subject: AgentEval, an open-source agent benchmarking harness
Today's date: May 18, 2026
The pitch: Subject line "Open-source eval framework for agentic AI — built after the Anthropic constitutional AI launch." It opens on Maxwell's Apr. 28 piece about developer adoption of Claude's agent SDK, then offers an open-source eval harness that benchmarks agent tool-use against his "real workflow" tests rather than synthetic ones. Repo opens Tuesday; first look plus two pilot customers (legal and healthcare) with workflow-time data; embargo through 9am ET Tuesday.

记者：Maxwell Zeff，TechCrunch
客户/主体：AgentEval，一款开源Agent基准测试工具
当前日期：2026年5月18日
稿件：主题行"Open-source eval framework for agentic AI — built after the Anthropic constitutional AI launch." 开篇提及Maxwell在4月28日发表的关于Claude的Agent SDK开发者采用情况的文章，随后介绍一款开源基准测试工具，可针对其"真实工作流"测试而非合成测试来评估Agent工具使用情况。代码库将于周二开放；提供提前查看权限及两位试点客户（法律和医疗领域）的工作流时间数据； embargo 至周二上午9点ET。

After

输出

Fit (confidence 0.86)

Why: Maxwell Zeff's Apr. 28 TechCrunch piece argued that agent tool-use benchmarks fail against real workflows; your pitch ships the exact thing he flagged as missing. The embargo, the repo opening, and the pilot customers give him a usable scoop.

Anchor article: "Developers say Claude's agent SDK is the easiest to deploy, hardest to evaluate" — TechCrunch, Apr. 28, 2026 — https://techcrunch.com/2026/04/28/claude-agent-sdk-eval-gap/. It names workflow-time benchmarks as the missing piece, which is exactly what the pitch launches.

What to do next: Nothing major — the target is right. Just normal copy polish.

Freshness: Most recent article May 16, 2026 (2 days ago). No freshness concern.

Where this came from: Web search. Checked the TechCrunch author page and recent results; the anchor link above supplied the match. Other recent bylines found May 3, May 11, and May 16.

Fit (置信度 0.86)

原因： Maxwell Zeff在2026年4月28日发表的TechCrunch文章指出，Agent工具使用基准测试无法匹配真实工作流；你的稿件恰好提供了他所提到的缺失解决方案。Embargo、代码库开放时间及试点客户为他提供了可用的独家内容。

锚定文章： "Developers say Claude's agent SDK is the easiest to deploy, hardest to evaluate" — TechCrunch, 2026年4月28日 — https://techcrunch.com/2026/04/28/claude-agent-sdk-eval-gap/。该文章指出工作流时间基准测试是缺失环节，而你的稿件正是针对此问题的解决方案。

下一步操作： 无需大幅修改——目标定位正确。仅需常规文案润色。

时效性： 最新文章发布于2026年5月16日（距今2天）。无时效性问题。

信息来源： 网页搜索。检查了TechCrunch作者主页及近期搜索结果；上述锚定链接提供了匹配依据。还找到了5月3日、5月11日和5月16日的其他近期署名文章。

Action

后续操作

Send the pitch after normal copy polish. The target is right.

完成常规文案润色后投递稿件。目标定位正确。

Example 2: No-Fit, Do Not Rescue It

示例2：不匹配，无需修改稿件

Before

输入

Journalist: Emily Chang (Bloomberg author page: https://www.bloomberg.com/authors/AS0E8x9aH-0w/emily-chang)
Today's date: May 18, 2026
The pitch: Subject line "New B2B accounting SaaS for freelancers." Two sentences: the team just launched an accounting platform for freelancers and would love coverage in Bloomberg.

记者：Emily Chang（彭博作者主页：https://www.bloomberg.com/authors/AS0E8x9aH-0w/emily-chang）
当前日期：2026年5月18日
稿件：主题行"New B2B accounting SaaS for freelancers." 内容为两句话：团队刚刚推出一款面向自由职业者的会计平台，希望获得彭博的报道。

After

输出

No-fit (confidence 0.41)

Why: Emily Chang's recent Bloomberg work is executive interviews, venture capital, and AI company strategy — not B2B SaaS launches or freelancer tooling. A freelancer-accounting product is outside her lane, and copy edits won't change that.

Anchor article: "Inside Anduril's pitch to the Pentagon" — Bloomberg, May 11, 2026 — https://www.bloomberg.com/news/articles/2026/05/11/anduril-pitch. It shows the company-and-executive-strategy level she covers, which contradicts the freelancer-SaaS ask.

What to do next: Nothing to fix — she's the wrong target. No edits, because the journalist is wrong, not the wording.

Freshness: Most recent article May 15, 2026 (3 days ago). No freshness concern.

Where this came from: Web search. Reviewed the Bloomberg author page and her last five visible articles; none touched SMB accounting, freelancer tooling, fintech for individuals, or product reviews.

No-fit (置信度 0.41)

原因： Emily Chang近期在彭博的作品聚焦高管访谈、风险投资和AI公司战略——而非B2B SaaS发布或自由职业者工具。面向自由职业者的会计产品不在其报道范围内，文案修改无法改变这一点。

锚定文章： "Inside Anduril's pitch to the Pentagon" — Bloomberg, 2026年5月11日 — https://www.bloomberg.com/news/articles/2026/05/11/anduril-pitch。该文章展示了她所覆盖的企业及高管战略层面内容，与自由职业者SaaS需求不符。

下一步操作： 无需修改——她不是合适的投递对象。请勿尝试为她修改稿件，因为是记者不合适，而非措辞问题。

时效性： 最新文章发布于2026年5月15日（距今3天）。无时效性问题。

信息来源： 网页搜索。查看了彭博作者主页及她最近五篇可见文章；均未涉及中小企业会计、自由职业者工具、个人金融科技或产品评测。

Action

后续操作

Drop the contact. Do not rewrite this for her.

放弃该联系人。请勿为她重写稿件。