Prompt Optimizer
v1.0.0Refine vague or poorly structured prompts into clear, structured, and effective prompts using proven prompt engineering techniques for improved AI output qua...
⭐ 0· 68·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (prompt refinement) aligns with the SKILL.md content: techniques, templates, and examples for improving prompts. It does not request unrelated credentials, binaries, or system access.
Instruction Scope
Runtime instructions are limited to analyzing and rewriting user prompts using the OPTIMIZE framework; they do not instruct the agent to read files, environment variables, network endpoints, or other system state outside of the provided user prompt.
Install Mechanism
No install spec and no code files — instruction-only. Nothing will be downloaded or written to disk by an installer.
Credentials
The skill declares no required environment variables, credentials, or config paths; the instructions likewise do not reference any secrets or external service tokens.
Persistence & Privilege
always:false (no forced inclusion). disable-model-invocation:false is the platform default and not itself a concern. The skill does not request persistent system changes or access to other skills' configuration.
Assessment
This is an instruction-only prompt-optimizing guide and appears internally consistent. Recommended precautions before installing or using it: (1) Do not paste secrets or sensitive data into prompts you ask the skill to optimize (the skill's scope is to rewrite whatever you give it). (2) Test optimized prompts on non-sensitive inputs to verify they produce the intended behavior and to detect undesired instructions or data exposure. (3) Be aware that the skill can be invoked by the agent (platform default); if you don't want autonomous use, restrict invocation in agent settings. (4) Remember optimized prompts can still cause downstream models to hallucinate or reveal data—validate outputs before sharing or taking action.Like a lobster shell, security has layers — review code before you run it.
aivk974g9nbsfw3ktda6p8920t8rh83q070latestvk974g9nbsfw3ktda6p8920t8rh83q070llmvk974g9nbsfw3ktda6p8920t8rh83q070optimizationvk974g9nbsfw3ktda6p8920t8rh83q070promptvk974g9nbsfw3ktda6p8920t8rh83q070prompt-engineeringvk974g9nbsfw3ktda6p8920t8rh83q070
License
MIT-0
Free to use, modify, and redistribute. No attribution required.