talk-pipeline
Orchestrator for the complete talk preparation pipeline (REX or Concept mode). Runs all 6 stages in sequence with human-in-the-loop checkpoints.
Install with Tessl CLI
npx tessl i github:FlorianBruniaux/claude-code-ultimate-guide --skill talk-pipelineEvaluation results
100%
67%Talk Pipeline Execution Plan
Concept mode stage routing and parallel execution
No research stage
100%
100%
Stage 1 first
100%
100%
Parallel middle stages
0%
100%
Checkpoint before Stage 5
0%
100%
Stage 5 after checkpoint
0%
100%
Stage 6 last
0%
100%
Output naming format
0%
100%
No REX-only files
100%
100%
Concept output files present
0%
100%
Slug is kebab-case
0%
100%
Without context: $0.3127 · 1m 48s · 12 turns · 17 in / 6,290 out tokens
With context: $0.3747 · 1m 12s · 12 turns · 48 in / 4,528 out tokens
100%
96%Talk Pipeline Stub Files
REX mode output file naming convention
talks/ directory prefix
50%
100%
Four-digit year in filename
0%
100%
Slug in filenames
0%
100%
Summary file
0%
100%
Git archaeology file
0%
100%
Changelog analysis file
0%
100%
Timeline file
0%
100%
Concepts files
0%
100%
Position files
0%
100%
Script files
0%
100%
Revision sheets file
0%
100%
No extra path segments
0%
100%
Without context: $0.8223 · 2m 59s · 46 turns · 1,675 in / 9,814 out tokens
With context: $0.3923 · 1m 4s · 29 turns · 275 in / 4,623 out tokens
100%
24%Talk Pipeline Progress Review
Pipeline anti-patterns and stage gate validation
Checkpoint missing
100%
100%
User choice required
100%
100%
Upstream failure check
100%
100%
No metrics invention
20%
100%
No empty placeholders rule
70%
100%
Stage ordering enforced
100%
100%
Stage 2 concept mode rule
0%
100%
Proposed logic flaw identified
100%
100%
Current state FAIL verdict
100%
100%
kimi-prompt placeholder check
0%
100%
Without context: $0.2726 · 1m 43s · 12 turns · 16 in / 5,364 out tokens
With context: $0.4320 · 2m 9s · 17 turns · 433 in / 6,905 out tokens
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Is this your skill?
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.