Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/mckruz/claude-code-mastery/performancegit clone --depth 1 https://github.com/MCKRUZ/claude-code-masteryWhat 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 | $0.00169 | $0.00169 |
| Opus 5 | $0.00084 | $0.00084 |
| Sonnet 5 | $0.00034 | $0.00034 |
| Haiku 4.5 | $0.00017 | $0.00017 |
Grade A, and why
performance 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 2d 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.
What it actually says
Performance
Context Window
- Avoid last 20% for large refactoring or multi-file features. Single-file edits and docs are fine near the limit.
- When context is getting long, finish current work cleanly rather than starting new complex tasks.
Research Time Limits
- If a single site or source takes more than 5 minutes to fetch or return useful results, abandon that source and move on.
- One or two attempts per source max, then pivot to the next.
Build Troubleshooting
- Build fails → analyze the actual error message first, don't guess.
- Fix incrementally — one error at a time, verify between fixes.
- Don't suppress warnings to make a build pass. Fix the underlying issue.
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.
- 2d ago First seen · 20 lines · 169 tokens per session scan A fe33ea166599
performance is a cursor rule published in the GitHub repository MCKRUZ/claude-code-mastery (11 stars, last pushed 4mo ago), licensed MIT. It adds 169 tokens to every session, about $0.0008 per session on Opus 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 2026-08-30.
Other cursor rules, from other repositories
cursorrules
Cursor rule "cursorrules" from syf2211/ruledoctor, covering 订单服务项目规则（demo） and 硬性规则.
integrations
External integration rules — providers, idempotency, bulkhead, observability.
decomposition
File and folder decomposition rules — when and how to split.
architecture
FastAPI layer architecture rules — Router → Service → Repository with Protocol injection.
llm-layer
LLM provider implementation patterns.
050-plan
When the user types /plan or asks to create a project plan, feature PRD, or retrospective.