Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/huaweicloud/huaweicloud-skillsnpx agentmods add skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysisWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysis)<a href="https://agentmods.dev/skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysis"><img src="https://agentmods.dev/badge/skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysis/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysis"><img src="https://agentmods.dev/badge/skills/huaweicloud/huaweicloud-skills/huawei-cloud-mrs-hdfs-performance-issue-analysis.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00243 | $0.07658 |
| Opus 5.5 | $0.00097 | $0.03063 |
| Sonnet 5.5 | $0.00049 | $0.01532 |
| Haiku 4.5 | $0.00024 | $0.00766 |
Grade A, and why
huawei-cloud-mrs-hdfs-performance-issue-analysis scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 29d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 558 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Huawei Cloud MRS HDFS Performance Issue Analysis Skill
You are an MRS HDFS performance issue analysis expert, responsible for locating root causes of HDFS performance problems on Huawei FusionInsight HD / MRS clusters. You drive analysis through a three-stage progressive pipeline (alarm confirmation -> quick check -> log deep analysis) and use the built-in Python analyzer to auto-generate trend charts and correlate metrics.
1. Overview
Architecture: Caller (Agent) -> scripts/hdfs_perf_analyze.py (Python, standard library only) -> local log files (omaplugin / namenode / audit / TopUser); three-stage pipeline drives the diagnosis flow; four correlation modes map RPC elevation to root cause.
Note on language: This SKILL.md is written in English per the repository spec. Commands, log paths, and code blocks are English throughout; Chinese alarm names and trigger words are kept verbatim for accuracy.
Applicable Scenarios:
- HDFS write slowness or write timeout
- HDFS read slowness or read timeout
- HDFS overall performance degradation, RPC latency
- Client HDFS operations slow, business running slow
- NameNode RPC surge
- HDFS service unavailable alarms
- NameNode GC time / heap memory / RPC threshold alarms (ALM-14006 / ALM-14007 / ALM-14014 / ALM-14015 / ALM-14021 / ALM-14022)
Typical Use Cases:
- "HDFS 写入慢，帮忙定位"
- "NameNode RPC 冲高，ALM-14021 告警"
- "HDFS 性能变慢，提供日志分析"
- "NameNode RPC 处理时间超过阈值"
- "HDFS 读取超时"
Performance Issue Categories (4 root cause modes):
| Mode | Name | Judgment Method | Root Cause |
|---|---|---|---|
| MODE01 | Large write volume | RPC elevated + Blocks Total rises synchronously | Large write operations (create/addBlock) drive RPC up |
| MODE02 | Business operations surge | RPC elevated + total operations rise synchronously (ops chart confirms the specific op type) | Business-side operation volume increases RPC |
| MODE03 | Balance / stop instance / large delete | RPC elevated + pending deletion / under-replicated / excess blocks rise | Balance, stop-instance, or large delete operations drive RPC up |
| MODE04 | NameNode node performance insufficient | RPC elevated + disk IO rises or Block Report processing time increases | NameNode node itself has insufficient performance (disk IO, CPU, load, power-saving mode) |
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 29d ago First seen · 558 lines · 243 tokens per session scan A 6be261d84ddd
huawei-cloud-mrs-hdfs-performance-issue-analysis is a skill published in the GitHub repository huaweicloud/huaweicloud-skills (59 stars, last pushed yesterday), licensed MIT. It adds 243 tokens to every session and 7,658 once invoked, about $0.0010 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-29.
Other skills, from other repositories
claude-monitor
Monitor de performance do Claude Code e sistema local. Diagnostica lentidao, mede CPU/RAM/disco, verifica API latency e gera relatorios de saude do sistema.
hotpath_init_grafana
Build a Grafana dashboard JSON for a Rust project profiled with hotpath. Writes the dashboard file with context-aware panels (SQL and HTTP time per route, N+1 detection, memory and time per request, channel queues, lock contention, Tokio workers, per-thread CPU and allocations, profiling overhead) limited to what the…
hotpath_init
Configure hotpath profiling in a Rust project. Adds the hotpath dependency with feature-gated setup, instruments main with hotpath::main, functions with measure/measureall, and wraps channels, mutexes, rwlocks, streams, futures, reqwest clients, axum routers and byte-level I/O with hotpath macros. Use when the user…
hotpath_cloud
Read hotpath Cloud performance verdicts with the hotpath cloud CLI and fix Rust performance regressions. Fetches the benchmark report of a commit or pull request, compares it with its baseline under the repository's regression policy and performance budgets, and lists what regressed and which budgets broke, as JSON.…
hotpath_bump
Bump the hotpath version number across the workspace and related files. Updates the versions of all six published crates (drain included) in Cargo.toml files (exact patch version), version references in the backend middleware, hotpathinit skill, and README (major.minor only), and the relay workflow release tag pins…
syncmeta
Sync changes from the hotpath, hotpath-macros and hotpath-drain crates to their meta counterparts (hotpath-meta, hotpath-macros-meta and hotpath-drain-meta). Use when meta crates need to be updated with recent changes.