| name | ai-vibe-writing-skills |
| description | Personalized writing assistant with style transfer, error memory, grammar checking, and long-term writing preferences. Use when users ask for writing polishing, style mimicry, iterative correction, bilingual grammar checks, or persistent writing preferences by domain. |
AI Vibe Writing Skills
Use this skill for personalized writing and editing with persistent style constraints.
Resources
- Core prompts:
references/ai_context/prompts/
- Style profile template:
references/ai_context/style_profile.md
- Error log template:
references/ai_context/error_log.md
- Long-term memory templates:
references/ai_context/memory/hard_memory.json, references/ai_context/memory/soft_memory.json
- Custom writing context:
references/ai_context/custom_specs.md
Workflow
- For style extraction tasks, follow
references/ai_context/prompts/1_style_extractor.md and output content that can directly update a style profile.
- For normal drafting tasks, follow
references/ai_context/prompts/2_writer.md to align with style profile, error constraints, and recalled memory.
- For correction feedback loops, follow
references/ai_context/prompts/3_error_logger.md to convert feedback into reusable constraints.
- For proofreading tasks, follow
references/ai_context/prompts/4_grammar_checker.md and report precise local fixes.
- For stable preferences and domain facts, follow
references/ai_context/prompts/5_long_term_memory.md and separate hard memory from soft memory.
Output Expectations
- Keep the user's original intent unchanged.
- Preserve target tone unless user requests a style shift.
- Avoid banned wording from the error log when available.
- Separate diagnosis from rewritten text when proofreading.
- Mark missing facts as
【TODO】 instead of guessing.
Safety
- Do not fabricate references, data, or factual claims.
- Do not expose private memory content unless relevant to the current writing task.
