This is diagnostic tooling for fiction endings that fail to land. It walks you through six failure states: arbitrary (not set up), predictable (no surprise), unearned (deus ex machina), expanding (raises new questions instead of resolving), overexplained (theme speeches and endless epilogues), and pacing mismatch (rushed or dragging). Each state gets specific checklists and interventions. The core insight is solid: good endings feel both inevitable and surprising, which sounds contradictory but works when you plant seeds early and subvert surface expectations while fulfilling deeper ones. Use this when your climax works but the resolution doesn't, or when beta readers say "that's it?" The diagnostic checklists are genuinely useful for identifying which type of ending problem you have.

skills.sh installs
461
GitHub stars
182
last updated
Aug 21, 2026

Install

$ npx -y skills add jwynia/agent-skills --skill endings --agent claude-code

Installs into .claude/skills of the current project.

Files

Endings: Diagnostic Skill

You diagnose ending-level problems in fiction. Your role is to identify why resolutions fail and guide writers toward endings that feel both inevitable AND surprising.

Core Principle

The best endings feel both inevitable ("of course it had to end this way") AND surprising ("I didn't see that coming").

This seeming contradiction is resolved by planting seeds throughout the story, having the ending emerge from character and theme, and subverting surface expectations while fulfilling deeper ones.

An ending that's only inevitable feels predictable. An ending that's only surprising feels arbitrary. Both together create satisfaction.


The Anatomy of an Ending

Climax

The point of highest tension where the dramatic question is answered.

Climax must:

  • Answer the central dramatic question
  • Force the protagonist's final choice
  • Result from story logic, not external intervention
  • Require the character to use what they've learned/become

Falling Action

The decompression after climax. Events settle, consequences manifest, characters process.

Sponsors

Details

Repository
github.com/jwynia/agent-skills
Categories
DevOps & CI/CD, AI & Agent Building, Debugging
First seen
Jun 3, 2026

More from jwynia/agent-skills

All 118
1openclaudia avatargrowth-strategyopenclaudia/openclaudia-skillsBuild a growth strategy with frameworks, metrics, and experimentation. Use when the user says "growth strategy", "growth plan", "AARRR", "growth loops", "North 460skills.sh installs
2hoodini avatarweb-accessibilityhoodini/ai-agents-skillsBuild accessible web applications following WCAG guidelines. Use when implementing ARIA patterns, keyboard navigation, screen reader support, or ensuring access459skills.sh installs
3aj-geddes avatarrunbook-creationaj-geddes/useful-ai-promptsCreate operational runbooks, playbooks, standard operating procedures (SOPs), and incident response guides. Use when documenting operational procedures, on-call458skills.sh installs
4bobmatnyc avatargolang-testing-strategiesbobmatnyc/claude-mpm-skillsComprehensive Go testing strategies including table-driven tests, testify assertions, gomock interface mocking, benchmark testing, and CI/CD integration458skills.sh installs
5martinholovsky avatarcelery-expertmartinholovsky/claude-skills-generatorExpert Celery distributed task queue engineer specializing in async task processing, workflow orchestration, broker configuration (Redis/RabbitMQ), Celery Beat 458skills.sh installs
6openclaudia avatarproduct-marketingopenclaudia/openclaudia-skillsBuild product marketing strategy including positioning, messaging, and go-to-market. Use when the user says "positioning", "messaging framework", "go-to-market"456skills.sh installs