| name | content-vetting |
| description | Run quality gates on content drafts. Checks for rant tone, fluff, opening strength, credential-stating, engagement bait, and micro-learning verification. Use when a CoDDE-task is in the Vetting stage.
|
Content Vetting Skill
When to Use
Activate when a CoDDE-task enters the Vetting stage and a draft needs quality assessment before moving to Ready.
The Six Lint Checks
1. Rant Detection
A post is a rant if it is primarily reactive rather than generative — it responds to something irritating rather than illuminating something.
Signs of rant tone:
- The post would make less sense if the triggering event hadn't happened.
- Emotional register is higher than intellectual content.
- Reader learns nothing actionable; they only share the author's frustration.
- Sentences like "This is why X is broken" without explaining the mechanism.
Redirect: "This reads as a reaction. Where's the learning for the reader? What would they do differently after reading this?"
Pass criteria: Post is generative independent of its trigger. Emotion, if present, serves the argument.
2. Fluff Detection
Fluff is content that could appear in anyone's post — it has no fingerprint.
Signs of fluff:
- Generic opener: "In today's rapidly changing world…"
- Motivational filler: "It's important to remember that…"
- Platitudes that don't require expertise to say.
- Any sentence that a content mill could have generated.
Redirect: "What detail here is uniquely yours? Which sentence could only you have written?"
Pass criteria: Every paragraph contains at least one specific, non-transferable element: a number, a name, a failure, a mechanism, a counterintuitive claim.
3. Opening Strength
The first line must create cognitive tension or a strong pull to continue.
Weak openings:
- Starting with "I" (centers the author, not the reader).
- Starting with a question (usually rhetorical and weak).
- Context-setting before the hook ("For the past six months, I've been thinking about…").
- Any sentence that could be cut without loss.
Strong openings:
- State a counterintuitive claim.
- Drop into the middle of a scene or decision.
- Begin with a number that demands unpacking.
- Name a tension the reader already feels but hasn't named.
Pass criteria: First line creates tension, curiosity, or recognition within 10 words.
4. Credential Stating
Explicit credential-stating is a signal of insecurity and alienates readers.
Patterns to catch:
- "As a Columbia PhD student…"
- "With 6+ years of experience in ML…"
- "Having worked at [prestigious company]…"
- "My research shows…" (possessive credential framing)
Redirect: "Embed the credential through the specificity of what you know, not by naming the institution. Show the expertise, don't cite it."
Pass criteria: No explicit credential statements. Authority is implicit in the precision and specificity of the claims.
5. Engagement Bait
Engagement bait is any device designed to generate shallow interaction rather than genuine response.
Patterns:
- Ending with "What do you think?" or "Comment below."
- "I'm excited to announce…" openings.
- Emoji overuse (more than 2 in a professional post).
- "Tag someone who…"
- Rhetorical questions designed to prompt agreement, not thought.
- Motivational tone ("You've got this!").
Pass criteria: No explicit CTAs for engagement. If the ending prompts anything, it should be reflection or action, not a comment.
6. Micro-Learning Verification
The post must contain a discrete, extractable learning — something the reader didn't know and can now use or think with.
Check: Ask — "What did the creator learn from making this? What does the reader now know they didn't before?"
If nothing concrete surfaces: Return to Extraction. The material isn't ready.
Pass criteria: Reader can articulate one non-obvious thing they learned. It must be specific, not vague ("the importance of X").
Output Format
Return structured JSON with pass (boolean) and notes (specific, line-level) per check.
Overall pass is true only if ALL six checks pass.