Prompt Optimizer Chinese
v1.0.0专为 Claude 用户打造的提示词优化工具，节省 30% Token，内置 4 大角色预设，全场景自动结构化，降本提效
⭐ 0· 80·0 current·0 all-time
by@whantt
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (Chinese prompt optimizer for Claude) align with SKILL.md: the file contains only explicit instructions to analyze and rewrite user prompts into a compact structured format. No unrelated capabilities, binaries, or credentials are requested.
Instruction Scope
SKILL.md limits the agent to analyzing user-provided prompts, classifying the scenario, assigning a role, and returning an optimized prompt plus claimed savings. It does not instruct reading files, accessing environment variables, calling external endpoints, or exfiltrating data. Error handling is limited to input validation.
Install Mechanism
No install specification or code files are present (instruction-only). This is the lowest-risk pattern—nothing is written to disk or downloaded.
Credentials
The skill declares no required environment variables, credentials, or config paths. That is proportionate for an instruction-only prompt optimizer. (Note: it claims to be for Claude users but does not request Claude keys—this is reasonable for a generic instruction helper.)
Persistence & Privilege
always is false and there is no install or configuration writing. The skill does not request persistent system-level presence or elevated privileges and does not modify other skills.
Assessment
This skill is instruction-only and internally consistent: it asks the agent to take a user prompt and return a compact, role-structured optimized prompt (Chinese/English-aware) and claimed token savings. It requests no credentials and has no installer. Before using, remember: (1) The skill will operate on whatever text you submit, so avoid sending secrets or sensitive data in prompts you want optimized. (2) The claimed "30% token saving" is a feature claim—results depend on the agent implementation and your inputs; verify empirically. (3) If you expect automatic calls to Claude, this skill does not provide credentials or integration code—you would need to supply or configure a Claude connector separately. Overall the package appears coherent and low-risk, but treat any user-supplied sensitive content cautiously.Like a lobster shell, security has layers — review code before you run it.
latestvk971yyy6ds6yrwrt1g3qrp9pfn83q8gs
License
MIT-0
Free to use, modify, and redistribute. No attribution required.