hop-planner
Between-hop reasoning for the research pipeline. Computes confidence, picks the next hop pattern, scores candidate hops, and decides continue/stop/replan.
Explore
119,435 skills indexed with the new KISS metadata standard.
Between-hop reasoning for the research pipeline. Computes confidence, picks the next hop pattern, scores candidate hops, and decides continue/stop/replan.
Fetches and summarizes one topic-hop's selected URLs by driving the researcher Python scripts (fetch_and_clean.py + summarize.py), returning per-source summaries. Used by the research-batch workflow, which has no Bash/Python of its own.
Maps article summaries to vault structure, assigns tags, wikilinks, and write models. Works with summaries instead of full content for token efficiency.
Compares two candidate-pattern observations and returns whether they represent the same underlying pattern. Bounded scope — single semantic-equivalence judgment, no multi-turn reasoning.
body payloads
logging
numerical value reversals
expected vs. anomalous behaviors
unhandled runtime exceptions causing service interruption)
Broken Object Level Authorization - BOLA)
not create an idealized study plan disconnected from reality. If 90+ requires luck
what wastes time)
rules
key shortcuts
deliver:**
why this duration
Mathematics
guys... we're all Fresh products. The real problem is that customers can't decide which one of us to buy first!
everyone runs to buy me!
but people run away from your heat!
I work all year long—winter
githubApp:
This is the reference for prompt engineering with Claude's latest models, including Claude Fable 5, Claude Mythos 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 4.6, and Claude Haiku 4.5. The page is organized in three parts:
Act as a technical documentation reviewer