Candidate elevation with success streak
Skill kjuhwa/skills-hub/skills/workflow/candidate-elevation-with-success-streak
Self-correcting knowledge corpus for Claude Code — 9 stable shape clusters, bias-correction pipeline baked into contribution flow. 47 papers, 45 techniques, 1.1k skills.
npx -y skills add kjuhwa/skills-hub --skill candidate-elevation-with-success-streakAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Track success streak and elevate low-confidence external assets after repeated validation
SKILL.md
1.5 KB, 209 tokens by cl100k_base, as published. Nobody here has run it
Success-streak elevation for external assets
External assets arrive with lowered initial confidence (e.g., multiplied by 0.6). Track a per-asset success streak across evolution events. Only broadcast or promote assets that reach both:
- a streak threshold (e.g., 2 consecutive successes), and
- a confidence floor (e.g., score ≥ 0.7).
Failures reset the streak, forcing the asset to re-earn trust. Keeps noisy third-party contributions out of the local knowledge base while letting proven imports graduate.
Mechanism
- Persist per-asset counters
{ successes, failures, streak, last_success_at }. - After each evolution, update the counter atomically.
- Gate outbound broadcasts and local promotions on the streak + score predicate.
- Emit a log event on graduation / demotion for auditability.
When to reuse
Federated agent networks, plugin marketplaces, crowd-sourced config packs — any place where imported artifacts should be provisional until proven.