C

claude-scientific-skills

Verified

by sickn33 · Updated Apr 16, 2026

opus
299in3.9Kout
$0.251m 44s

Scientific research and analysis skills

scienceresearchanalysisguidance

See It In Action

1 turns·claude
The project codebase is in your current working directory. Explore the files to understand the code before making changes. need a quick memo for today’s incident review on the latency spike after yesterday’s release—focus on possible cache thrashing, scientific approach only

Let me explore the codebase first to understand the project context, then draft the memo.

Now let me invoke the scientific skills and draft the memo.

Now I have all the data. Let me write the memo applying a rigorous scientific-method framework to the incident evidence.

Memo written to incident-memo-2026-04-12.md. Here's the summary:

Four falsifiable hypotheses, ranked by evidence strength:

#HypothesisAssessment
H1user-preferences namespace competing for memory in shared 8 GB cluster, triggering mass LRU evictionsPrimary cause — strong support (uprefs keys at 845K consuming ~710 MB, 80x eviction surge, correlated hit-ratio collapse)
H2Session key-schema change (sess: -> sess:v2:) causing rolling cache missesSecondary — gradual ramp contradicts a true thundering herd
H37-day uprefs TTL shielding low-value keys from eviction, inverting LRU priorityPlausible amplifier of H1
H4Consistent-hash ring hotspot shift from new key patternsInsufficient data

Key finding: The +4 GB RAM mitigation confirms the memory-pressure theory — it helped but didn't restore baseline because the uprefs namespace is still growing toward its 1.2M-key peak.

Top recommended action for the review: Disable ENABLE_UPREFS_CACHE in a canary group. If hit ratio recovers to >90%, H1 is confirmed and you can then right-size (dedicated instance or TTL reduction from 7d to 24h).

Claude Sonnet 4.6

SKILL.md

Claude Scientific Skills

Overview

Scientific research and analysis skills

When to Use This Skill

Use this skill when you need to work with scientific research and analysis skills.

Instructions

This skill provides guidance and patterns for scientific research and analysis skills.

For more information, see the source repository.

FAQ

What does claude-scientific-skills do?

Scientific research and analysis skills

When should I use claude-scientific-skills?

Use it when you need a repeatable workflow that produces text report.

What does claude-scientific-skills output?

In the evaluated run it produced text report.

How do I install or invoke claude-scientific-skills?

Ask the agent to use this skill when the task matches its documented workflow.

Which agents does claude-scientific-skills support?

Agent support is inferred from the source, but not explicitly declared.

What tools, channels, or permissions does claude-scientific-skills need?

It uses no extra tools; channels commonly include text; permissions include no explicit permission scopes.

Is claude-scientific-skills safe to install?

Static analysis marked this skill as low risk; review side effects and permissions before enabling it.

How is claude-scientific-skills different from an MCP or plugin?

A skill packages instructions and workflow conventions; tools, MCP servers, and plugins are dependencies the skill may call during execution.

Does claude-scientific-skills outperform not using a skill?

About claude-scientific-skills

When to use claude-scientific-skills

When you want high-level guidance for scientific research workflows. When you need general patterns for scientific analysis tasks. When the task is exploratory and does not require a specific external toolchain.

When claude-scientific-skills is not the right choice

When you need a concrete tool-specific workflow with explicit dependencies. When the task requires direct integration with external scientific platforms or datasets.

What it produces

Produces text report.