一键导入
spec-navigator
Efficiently retrieve specific sections from requirement documents (prompts) to save context tokens.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Efficiently retrieve specific sections from requirement documents (prompts) to save context tokens.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | spec_navigator |
| description | Efficiently retrieve specific sections from requirement documents (prompts) to save context tokens. |
This skill allows you to query specific sections of the master instruction file (inst.md) or any markdown file, saving you from reading the entire file into the context window of the LLM.
To significantly reduce token consumption when working with large specification files like inst.md by only loading the relevant section (e.g., "Architecture", "Stack", "MVP") into the conversation context.
scripts/query_md_section.pyExtracts a section and its subsections from a markdown file based on a header search term.
Usage:
python3 .agent/skills/spec_navigator/scripts/query_md_section.py [ABSOLUTE_PATH_TO_FILE] [HEADER_SEARCH_TERM]
Workflow Example:
inst.md, run:
python3 .agent/skills/spec_navigator/scripts/query_md_section.py /home/crrb/ag_projects/portal_da-tica/inst.md "Persistencia"
inst.md"Arquitectura" - For stack and module layout."MVP" or "Alcance" - For the immediate to-do list."Persistencia" - For database and volume mounts."Restricciones" - For negative constraints (what NOT to do).