Agent Skills › ory/lumen › reindex

reindex

GitHub

提供Lumen项目索引的刷新或重建服务。支持通过MCP工具自动更新，或根据用户请求执行CLI强制重建、清理过期索引等操作，并报告重建后的索引状态与统计信息。

plugins/lumen/skills/reindex/SKILL.md ory/lumen

Trigger Scenarios

用户要求刷新或重新构建项目索引用户询问索引状态或需要清理缓存

Install

npx skills add ory/lumen --skill reindex -g -y

More Options

Non-standard path

npx skills add https://github.com/ory/lumen/tree/main/plugins/lumen/skills/reindex -g -y

Use without installing

npx skills use ory/lumen@reindex

指定 Agent (Claude Code)

npx skills add ory/lumen --skill reindex -a claude-code -g -y

安装 repo 全部 skill

npx skills add ory/lumen --all -g -y

预览 repo 内 skill

npx skills add ory/lumen --list

SKILL.md

Frontmatter

{
    "name": "reindex",
    "description": "Refresh or rebuild the bundled Lumen index for the current project, preferring MCP-driven refreshes and using the CLI only for an explicit clean rebuild."
}

Lumen Reindex

Refresh or rebuild the bundled Lumen index for the current project.

Steps

Call the Lumen index_status tool for the current working directory so you can report the current state before making changes.
If the user wants the index refreshed or seeded, call the Lumen semantic_search tool with a broad natural-language query and set path or cwd to the current working directory. The search tool refreshes stale or missing indexes automatically.
If the user explicitly asks for a clean rebuild, explain the options and run one via the shell:
- lumen index --force . — reprocesses every file for the current project without wiping other worktrees or their shared vectors. Prefer this.
- lumen clean --days 0 && lumen index . — deletes every cached index not held by an active indexer lock before rebuilding; locked collections are skipped and counted in the summary. Use only when the user asks for a full wipe.
- lumen clean — removes indexes for projects that no longer exist or have not been used in 30 days. Use to reclaim disk space, not to rebuild the current project.
After the refresh or rebuild, report the new index status, including vector precision, unique vectors, shared references, deduplication ratio, database bytes, and reclaimable bytes when available.