repurpose-content — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited repurpose-content (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.
One strong piece of source content can become five to ten LinkedIn posts, each from a different angle, because long-form material almost always contains multiple distinct ideas the original author treated as one. This skill exists to find those distinct ideas and shape each into something LinkedIn-sized, in the user's voice, without producing variations on the same thing.
Most LinkedIn AI tools that claim "repurposing" just rewrite the source into a shorter version, then rewrite it again with different phrasing. The result is five posts that say the same thing five ways. They cannibalize each other and bore the audience.
Real repurposing finds the different claims, examples, and moments inside the source and shapes each one into its own LinkedIn-native piece. A talk has multiple beats; each beat can be its own post. A blog post has a main argument and several supporting points; the supporting points can be standalone posts. A podcast transcript has the speaker saying surprising things at specific moments; those moments become posts.
The test: if two of the repurposed pieces could be confused for each other, the repurposing failed.
Trigger when:
Do NOT trigger when:
The source content. Either pasted into chat, uploaded as a file, or linked (the user needs to paste the text since URLs often will not fetch). Plus:
If voice-profile.md exists, read it.
A good repurposing pass has three steps:
1. Read the source for distinct ideas. A 30-minute talk has roughly 4-8 distinct beats. A 1,500-word blog post has 3-5. A podcast episode usually has 5-10 "moments" (a surprising claim, a specific story, a turn of phrase). The first job is to identify these without summarizing the whole.
2. For each distinct idea, identify its anchor. What is the specific moment, quote, number, or detail from the source that makes this idea LinkedIn-postable? If you cannot point to a specific anchor in the source, the idea is too generic.
3. Choose a format and an angle. Each piece gets:
The output of this skill is essentially a batch of find-ideas seeds, each derived from the source, ready to be sent to the appropriate writing skill.
Source: [type and brief description]
Pieces produced: [N]
---
Piece 1: [format]. [one-line angle]
Source anchor: [the specific line, quote, moment, or detail from the source that grounds this piece]
Why this angle: [one sentence on what makes it postable]
Best fit for: [write-post / build-carousel / write-video-script / write-longform]
---
Piece 2: [format]. [one-line angle]
Source anchor: ...
[continue for each piece]
---
Suggested cadence: [if relevant: how to space the pieces over time so they do not feel like a barrage from one source]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.
Repurposing produces multiple pieces from one source. Posting them all in one week reads as "I had one thought and milked it." Space them. A reasonable default: one piece per week, two weeks between the strongest two pieces. If the source is from a talk the user gave, the first piece can reference the talk explicitly; the later ones should stand alone without that frame.
For newsletter writers, the source may be the basis of one full issue plus 2-3 feed posts that point back to the newsletter.
Source: A 45-minute podcast transcript where the user (a CTO) talks about scaling an engineering team from 5 to 30 people.
User asks: "Give me 5 LinkedIn pieces from this."
Output:
Source: 45-min podcast on scaling an eng team from 5 to 30
Pieces produced: 5
---
Piece 1 (post): The first hire who is not technical breaks the team's mental model, and nobody warns you
Source anchor: At ~12:00, you say "the day we hired our first PM, every engineer's calendar changed and we had no idea why."
Why this angle: A specific, hard-to-fabricate moment with a concrete trigger (the PM hire) and a non-obvious consequence (calendar disruption). Almost certainly resonates with anyone who has scaled past 10.
Best fit for: write-post
---
Piece 2 (carousel): How our 1:1 system broke at every team size, and what we changed each time
Source anchor: At ~24:00, you describe the 1:1 cadence at 5, 10, 20, and 30 people and what failed at each step.
Why this angle: Has natural sequential structure (4 size milestones, 4 failure modes, 4 fixes). Perfect carousel content.
Best fit for: build-carousel
---
Piece 3 (post): I stopped doing technical interviews when the team hit 15
Source anchor: At ~31:00, the moment you describe handing off the technical bar to a hiring committee.
Why this angle: A specific change of mind by a CTO, with a clear trigger (size 15). The kind of contrarian claim that earns comments from people on both sides of the founder/IC debate.
Best fit for: write-post
---
Piece 4 (video script): The metric I track instead of velocity at this stage
Source anchor: At ~38:00, you say "velocity stopped meaning anything around size 20. I started tracking how often someone says 'I don't know who owns this.'"
Why this angle: The specific surrogate metric is unusual and memorable. Works well on camera because the surprise is in the speaker's voice.
Best fit for: write-video-script
---
Piece 5 (longform): What "founder mode" actually breaks at size 25
Source anchor: A throughline across the second half of the transcript, where you push back on the popular "founder mode" framing.
Why this angle: A point of view on a current debate, with first-hand evidence at a specific scale. Long-form because the argument needs space to engage with the opposing position.
Best fit for: write-longform
---
Suggested cadence: Lead with Piece 3 (the change of mind), follow two weeks later with Piece 1 (the PM moment), then Piece 4 (video) the following week. Save Piece 2 (carousel) for a week when you have time to design slides. Piece 5 (longform) when you have a Saturday morning.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.