Compress prompts with lossy or lossless modes while retaining directives and constraints.
npx skills add https://github.com/jbrukh/skills --skill compress-prompt
Please help me install this Agent Skill. Skill: compress-prompt Source: https://github.com/jbrukh/skills/tree/main/skills/compress-prompt Command: npx skills add https://github.com/jbrukh/skills --skill compress-prompt
Compress prompts while preserving semantic content, reducing length while keeping all instructions, constraints, and essential structure intact, enabling faster iteration and shareability.
Provide a prompt and specify a mode (--lossless for full retention) to receive a compressed version with a concise stats report.
High-intent search queries and answers about installing and using this skill.
You can compress prompts while maintaining semantic fidelity by using lossy or lossless modes, targeting 30-50% or 10-30% length reduction respectively. This preserves instructions and constraints for faster iteration.
Yes, you can reduce system prompt length while keeping all directives intact by using the lossless mode. It enforces explicit retention of all constraints and achieves a 10-30% reduction without losing essential instructions.
Lossy prompt compression targets 30-50% reduction with at least 90% semantic retention, while lossless compression preserves 100% of directives and constraints with a smaller 10-30% reduction.
Yes, prompt compression works on multi-section task instructions and recovers intent even when formatting is complex. It applies to any prompt type including system prompts and multi-section documents.
You verify that your compressed prompt retains original intent through a built-in validation workflow that checks target reductions and enforces explicit retention of all directives and constraints.
You should avoid lossy prompt compression when your task requires 100% preservation of all directives and constraints, opting instead for lossless compression to ensure no critical instructions are dropped.
thejefflarsonDetect and fix prompt construction patterns that allow user input to influence system prompts.
kinnerkarmanishOptimize prompts through an iterative meta-prompt loop with evaluation metrics.
bmad-code-orgAudit translated documentation against English source files for fidelity.
burgebjCreate structured prompts with roles, constraints, and verification steps.
