Stage 4 — Strategic angles, titles, descriptions, peer feedback draft. Includes mandatory CHECKPOINT before script can start.
Install with Tessl CLI
npx tessl i github:FlorianBruniaux/claude-code-ultimate-guide --skill talk-stage4-positionAngle generation and recommendation structure
Angle count in range
100%
100%
Angles genuinely distinct
100%
100%
Audience fit per angle
0%
87%
Strengths and weaknesses per angle
100%
100%
Verdict star rating per angle
0%
100%
Clear argued recommendation
100%
100%
Recommendation not last-angle bias
100%
100%
Integration justification
100%
100%
Narrative arc justification
100%
100%
Duration sustainability check
100%
100%
Recommended structure table
50%
100%
Goal and audience header
50%
100%
Without context: $0.1574 · 1m 10s · 8 turns · 13 in / 3,176 out tokens
With context: $0.8177 · 4m 21s · 28 turns · 33 in / 13,032 out tokens
Title options and description quality
Title count for main angle
100%
100%
Backup titles for other angles
0%
100%
Title verdict/recommendation
100%
100%
No clickbait titles
100%
100%
Short description word count
0%
100%
Long description word count
0%
100%
Description opening style
100%
100%
No marketing language
100%
100%
Not a slide summary
87%
100%
Speaker pitch present
87%
100%
Tags/keywords present
100%
75%
Long description includes metrics
100%
100%
Without context: $0.2880 · 1m 53s · 13 turns · 20 in / 5,099 out tokens
With context: $0.7118 · 3m 10s · 27 turns · 1,047 in / 9,881 out tokens
Full pipeline output, CHECKPOINT, and feedback draft
angles.md file created
100%
100%
titre.md file created
100%
100%
descriptions.md file created
100%
100%
feedback-draft.md file created
100%
100%
CHECKPOINT message saved
100%
100%
CHECKPOINT lists all 4 files
100%
100%
CHECKPOINT asks angle choice
100%
100%
CHECKPOINT asks title choice
100%
100%
Feedback draft DM format
0%
100%
Feedback draft email format
100%
100%
Feedback draft public format
0%
100%
Pre-send checklist in feedback draft
25%
100%
No Stage 5 invocation
100%
100%
Without context: $0.6744 · 4m 10s · 21 turns · 26 in / 13,406 out tokens
With context: $0.8619 · 3m 29s · 35 turns · 225 in / 11,296 out tokens
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.