write-longform — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited write-longform (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
LinkedIn long-form content is genuinely different from a 1,300-character feed post. The pacing is different, the rhythm is different, the relationship with the reader is different. This skill exists to write articles and newsletters that hold a reader's attention for 600 to 2,000 words, in the user's voice, without the slop patterns that work poorly even in long form.
LinkedIn is investing heavily in long-form content. Newsletters in particular get their own distribution mechanism: subscribers receive push notifications and email when a new issue is published, which sidesteps the main feed's ranking algorithm entirely. Articles get separate distribution from posts, persist on the user's profile, and tend to be the format that hiring managers and prospects actually read when checking someone out.
But long-form is harder than short-form. A weak feed post is forgettable. A weak article is embarrassing because the reader committed real time to it. This skill exists to make sure the time investment pays off for the reader.
Trigger when:
Do NOT trigger when:
If voice-profile.md exists, read it.
A LinkedIn article or newsletter that works has four traits:
1. A real claim, not a survey. The piece argues something. "Five trends in B2B SaaS" is not a piece; it is a list of trends. "B2B SaaS is undervaluing customer success because it confuses retention with renewal" is a piece. The claim should be specific enough that someone could disagree with it.
2. Earns the time investment. Every paragraph should be earning the reader's next paragraph. If a paragraph could be cut without losing the argument, cut it. Length is not a virtue.
3. Specific, anchored evidence. Numbers, examples, named cases, quotes. Long-form fails when it tries to substitute eloquence for evidence. The reader committed time; they want to come away with something concrete.
4. A point of view, not a survey of points of view. "On one hand X, on the other hand Y, the truth is probably somewhere in between" is the worst possible long-form pattern. Take a side. Acknowledge the opposing view in one paragraph and then defend yours.
Long-form benefits from sub-headers more than feed posts do. The reader can navigate. The piece feels less like a wall of text. Use 3-5 sub-headers for an 800-word piece, 5-8 for a 1,500-word piece.
A common structure that works (do not treat as a template; the piece should serve the argument):
Newsletters have subscribers who chose to receive them. This changes the writing in three ways:
For a first newsletter issue, set the premise. What is the newsletter about, how often does it publish, what kind of pieces should the reader expect.
Output must read as written by a human on the first pass. The constraints below are how you write, not a checklist to apply later. Do not narrate this process to the user. Do not show a "before humanizer / after humanizer" sequence. Just produce clean output.
Hard bans — never appear in output:
- **Speed:** ...). Write as prose.Required patterns:
Final scan before output:
— and –. Any hit means redraft.Full 33-pattern reference: see humanizer/SKILL.md in this plugin.
Return the article as markdown, with:
Angle: B2B SaaS undervalues customer success because the team measures renewal, not retention. The two are different.
User confirms: "Make this an article, ~1,000 words."
Output (truncated):
# B2B SaaS is measuring the wrong thing about customer success
The team I joined last year had a 94% renewal rate and a customer success function in serious trouble. Renewal rates were green. Customer feedback was red. Two of our largest accounts were planning to leave at the next renewal window and had told their account managers. The board did not know.
This is the gap I want to write about. Renewal is what shows up on the dashboard. Retention is what the customer is actually doing. They look the same on a one-year contract. They diverge sharply on a three-year one. Almost every CS team I have worked with measures the first and assumes it tells them about the second.
## The argument I want to engage with first
The strongest case for measuring renewal is that it is the only metric that matters at the contract boundary. A customer who renews paid the bill. A customer who stays engaged but does not renew is a story about engagement, not revenue. From a finance standpoint, renewal is the truth.
This argument is correct as far as it goes. The problem is that it assumes renewal is the leading indicator. It is not. By the time renewal happens, the decision was made months ago, often a year ago. Measuring renewal is measuring an outcome you cannot influence.
## What retention actually looks like
[continues for ~600 more words, with named examples and specific numbers]
## Where I might be wrong
The case against retention metrics is that they are noisy. A customer's product usage can drop because their team is reorganizing, not because they are leaving. A churned account can look healthy on retention metrics right up until the cancellation email. This is real. But the answer is not to give up on retention; it is to instrument it better.
## What follows from this
If you run a CS team, you probably already track usage. The question is whether usage is a top-line metric or a debugging tool. Make it top-line. Put it next to renewal in your weekly review. Investigate the accounts where the two metrics diverge.
The accounts where they diverge are the accounts that will surprise you in 18 months. That is what I wish I had known a year ago.(~1,050 words total. Article argues a specific claim, engages the opposing view in section two, anchors with the 94%/board story, closes with an actionable implication.)
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.