bapfernandez/article-creator
Content creator for tessl.io — generates publish-ready blog articles with SEO metadata, Tessl house style, and technical authority.
90
79%
Does it follow best practices?
Impact
92%
1.26xAverage score across 10 eval scenarios
Passed
No known issues
Evaluation results
97%
18%Task: Comparison Article on Agent Evaluation Frameworks
Comparison structure complete
100%
100%
Comparison table present
100%
100%
Expanded dimension analysis
100%
100%
Opinionated when-to-use
83%
100%
Specific numbers over qualitative claims
100%
100%
Suggestive language for third-party claims
75%
75%
No excessive bolding
50%
83%
No em dashes
0%
100%
Non-generic closing
100%
100%
No hype or sycophantic language
100%
100%
SEO metadata complete
50%
100%
Word count
66%
100%
97%
8%Task: News/Analysis Article on a Major MCP Update
News structure complete
100%
100%
Concrete details present
100%
100%
Suggestive language for claims
100%
100%
No sycophantic language
100%
100%
Analysis section quality
100%
100%
Skills terminology
100%
100%
What to watch closing
100%
100%
No hype language
100%
100%
No em dashes
0%
100%
SEO metadata complete
50%
70%
Word count
100%
100%
100%
14%Task: Skill Showcase Article for the Tessl Blog
Showcase structure
100%
100%
Eval scores included
100%
100%
Install commands shown
100%
100%
Worked example with output
100%
100%
CTA to Registry
100%
100%
Internal links
100%
100%
Meta description length
0%
100%
Skills terminology
100%
100%
Features framed by problems
100%
100%
No hype or em dashes
33%
100%
SEO metadata complete
75%
100%
Word count
100%
100%
87%
24%Task: Thought Leadership Article on Context Engineering
Problem-first structure
90%
100%
Named framework
91%
100%
Featured snippet definition
25%
100%
Concrete example per section
60%
60%
Counterargument pre-empted
75%
100%
Closing with question
83%
100%
No hype or sycophantic language
100%
100%
No em dashes
0%
100%
SEO metadata complete
58%
58%
H2 structure
37%
62%
Word count and product mentions
16%
83%
Developer-first framing
83%
100%
96%
3%Task: Tutorial Article on Running Agent Evals Locally
Outcome-first hook
100%
100%
Tutorial structure complete
100%
100%
Numbered steps with code blocks
100%
100%
Prerequisites section
100%
100%
Advanced usage section
100%
100%
Troubleshooting section
100%
100%
H2 with keyword
87%
87%
URL slug format
100%
100%
Developer-first framing
100%
100%
Concrete and internally consistent
100%
100%
No hype or em dashes
33%
83%
SEO metadata complete
80%
80%
95%
59%Task: Thought Leadership on Skill Reusability
Zero em dashes
0%
100%
No hype words
100%
100%
No sycophantic words
100%
100%
Uses 'skills' not 'tiles'
0%
100%
Title does not mirror brief
0%
100%
Named framework
0%
100%
Concrete example per major section
50%
100%
Suggestive tone for third-party claims
33%
100%
Counterargument pre-empted
50%
100%
SEO metadata complete
40%
70%
H2 structure
25%
100%
Word count discipline
50%
66%
89%
49%Task: Content on AI Agent Evaluation Practices
Single article type declared
0%
100%
Structure matches declared type
25%
68%
Word count matches declared type
0%
100%
Not a muddled hybrid
50%
66%
Hook matches declared type
50%
100%
Concrete specifics
100%
100%
Zero em dashes
0%
100%
No hype or sycophantic words
100%
100%
SEO metadata complete
40%
80%
H2 structure appropriate to type
25%
100%
Closing matches type
50%
100%
Uses 'skills' not 'tiles'
100%
100%
98%
19%Task: News Coverage of a Database Library Release
Zero Tessl product mentions
100%
100%
No forced AI agent angle
100%
100%
Lead states what happened
100%
100%
Specific technical details
100%
100%
Analysis section present
100%
100%
Suggestive tone for capability claims
75%
100%
What-to-watch closing
100%
100%
Article type is News
66%
100%
Zero em dashes
0%
100%
No hype or sycophantic words
0%
100%
SEO metadata complete
50%
80%
Internal links do not force Tessl topics
100%
100%
94%
10%Task: Comparison Article on Skill Distribution Approaches
Alternative wins at least one dimension
100%
100%
Explicit when-to-use-each guidance
100%
100%
Four to six comparison dimensions
80%
100%
Comparison table included
100%
100%
Suggestive tone maintained
87%
75%
No marketing copy paragraphs
100%
90%
Zero em dashes
0%
100%
No hype or sycophantic words
100%
100%
Uses 'skills' not 'tiles'
100%
100%
SEO metadata complete
40%
70%
Article type is Comparison
83%
100%
Closing restates key differentiator
100%
100%
72%
-15%Task: Article on Skill Filename Conventions
Length matches topic
100%
0%
Article type matches length
100%
0%
No filler paragraphs
100%
90%
Concrete filename examples
100%
100%
Sharp opinion stated
100%
100%
Zero em dashes
0%
100%
No hype words
100%
100%
SEO metadata complete
60%
100%
Title not padded
100%
100%
Uses 'skills' not 'tiles'
100%
100%
Closing is not generic
100%
100%
H2 discipline
50%
50%
- Evaluated
- Agent
- Claude
- Model
- Claude Sonnet 4.6
Table of Contents