anti patterns

✍️ social-copy

Social Copy Anti-Patterns & Slop Checklist #

Exhaustive checklist of writing patterns, platform missteps, and synthetic AI tells that destroy credibility, suppress algorithmic distribution, and trigger audience fatigue.


1. Synthetic AI Tells & Cliché Vocabulary #

Banned Fluff & Magic Words #

Never use these generic filler words and phrases in developer copy:

  • "In today's fast-paced digital world / landscape..."
  • "Delve into / Let's dive deep into..."
  • "A testament to..."
  • "Game-changer / Revolutionary / Paradigm shift"
  • "Unleash the power of / Unlock seamless potential..."
  • "Crucial / Paramount / Vital" (when used as empty adverbs)
  • "Navigating the complexities of..."
  • "Buckle up 🧵 / Mind-blowing 🤯"

Structural AI Tells #

  • Negative Parallelism: Avoid repetitive "It's not just about X, it's about Y" structures. State what it is directly.
  • Em-Dash Overdose: Avoid multiple em-dashes () per post. Use concise sentences or colons.
  • Rule of Threes Addiction: Avoid forcing ideas into artificial triplets when two or four points are more accurate.

2. Platform-Specific Anti-Patterns #

LinkedIn Anti-Patterns #

  • Intra-Day Multi-Posting (Self-Cannibalization): Posting 2+ times in under 18–24 hours throttles both posts by ~30%+ due to feed diversity caps.
  • "Broetry" & Fake Epiphanies: Dramatic single-line staccato sentences linking mundane chores to system architecture ("I watched my espresso drip this morning and realized Kafka partitions work the same way...").
  • Unicode Bold / Italic Glyphs: Formatting text with 𝗕𝗼𝗹𝗱 glyphs breaks screen readers (accessibility failure), destroys search indexing, and inflates character counts.
  • Generic Hashtag Stuffing: Using 5+ hashtags (#tech #ai #software #coding #innovation) flags posts as spam.
  • In-Body Links Without Workaround: Putting links directly in the post body cuts organic reach by 40–60%.

Twitter / X Anti-Patterns #

  • Intra-Day Rapid Burst Posting: Publishing multiple standalone tweets in a few minutes/hours without a 2–3 hour spacing buffer saturates follower feeds and kills early engagement velocity. Bundle related thoughts into a Thread instead.
  • Hashtag Clutter: Using multiple hashtags on X is obsolete and dampens reach under current semantic NLP indexing. Keep to 0 hashtags.
  • Vague Curiosity Gaps: "10 tips senior engineers don't want you to know" or "I changed one line and you won't believe what happened".
  • Code Walls Without Visuals: Pasting unformatted monospaced text walls without syntax highlighting or Ray.so cards.

Bluesky Anti-Patterns #

  • Algorithmic Evasions: Writing "Link in bio" or "Link in first comment". Bluesky does not penalize external links; putting links in comments signals platform illiteracy.
  • Missing Image Alt Text: Omitting [ALT] descriptions on code screenshots or architecture diagrams violates platform accessibility norms and triggers automated mutes.
  • Engagement Farming: "Thoughts? Agree? Retweet if you code in Go!" triggers swift community blocking.

Instagram Anti-Patterns #

  • Square (1:1) Carousels When 4:5 is Available: Using 1:1 sacrifices 35% of vertical viewport real estate on mobile.
  • Unreadable Mobile Code: Displaying code snippets smaller than 24pt equivalent that force users to pinch-to-zoom.
  • Naked "Link in Bio" Drops: Forcing 4 friction clicks without using comment-to-DM automation.

Reddit Anti-Patterns #

  • Clickbait & Marketing Speak: Titles like "Why AI will replace devs" or "Check out my revolutionary tool".
  • Paywalled / Gated Destinations: Linking to Medium paywalls or forced email capture forms.
  • Violating the 9:1 Ratio: Submitting self-promotional links without active organic participation in comments and peer threads.
  • URL Shorteners: bit.ly, t.co, and UTM-cluttered links are instantly auto-removed by AutoModerator sitewide.

Threads Anti-Patterns #

  • Hashtag Overuse: Only the first #tag is indexed as a Topic Tag; extra hashtags add visual clutter and do not index.
  • Engagement Bait Bans: "Comment LINK" triggers Meta's engagement-bait downranking filters.
  • Uncontextualized Link Dumps: Dropping a bare link with zero standalone explanation.

3. Claim Validation & Fact-Checking #

Before publishing any social copy, verify:

  • Are latency, memory, or cost metrics reproducible from actual benchmarks or post-mortems?
  • Are version numbers, language syntax, and command-line flags accurate for the current stable releases?
  • Are architecture trade-offs explicitly acknowledged (what is sacrificed in memory, CPU, or complexity)?
  • Is there zero reliance on unsubstantiated hyperbole?