l-pedia-search

Search and read Confluence L-Pedia documentation to produce cited case briefs.

Updated Aug 10, 2026
One-click install
npx skills add https://github.com/neuthos/ldx-monorepo --skill l-pedia-search-neuthos
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: l-pedia-search
Source: https://github.com/neuthos/ldx-monorepo/tree/main/.agents/skills/l-pedia-search
Command: npx skills add https://github.com/neuthos/ldx-monorepo --skill l-pedia-search-neuthos

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Answers L-DX case questions about screens, business flows, and Ringi decisions by grounding them in the L-Pedia Confluence space, which is auth-gated and unreachable by generic web readers. ## Core Features & Use Cases - Confluence MCP Search: Uses the read-only confluence MCP server (CQL search, page tree, page fetch) with a ./scripts/lpedia CLI fallback for terminal use and query debugging. - Scoped Discovery: Locates the EN/Manuals and EN/Ringi roots each session and narrows searches with ancestor scoping, supporting both English and Japanese keywords. - Cited Case Briefs: Produces an answer, a sources table with URLs, verbatim evidence excerpts (Japanese terms preserved with English glosses), and explicit gaps. - Use Case: Ask how Store Sales cancellation works; the skill searches L-Pedia, reads the relevant manual and Ringi pages, and returns a cited brief listing what is and is not documented. ## Quick Start Ask how a specific L-DX screen or business flow works and request a cited answer from the L-Pedia Confluence documentation.

Frequently Asked Questions about l-pedia-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search Confluence documentation from an AI agent?▼

Use the read-only confluence MCP server tools such as confluence_search with CQL queries, confluence_get_page, and confluence_get_page_children. A CLI fallback script provides the same search, tree, and page operations for terminal use.

How to write CQL queries scoped to a Confluence page tree?▼

Start with space = "LPedia" AND text ~ "keywords", then narrow with AND ancestor = "<root-page-id>" to scope results under a specific tree like Manuals or Ringi. Use title ~ when hunting a page by name.

Why does Confluence return 404 for a space that exists?▼

Confluence returns 404 both for a wrong space key and for an account lacking permission to see the space. On auth-shaped 404s, verify the CONFLUENCE_URL, CONFLUENCE_USERNAME, and CONFLUENCE_API_TOKEN values in .env before assuming the key is wrong.

Can I use WebFetch to read Atlassian Confluence pages?▼

No. Auth-gated Confluence spaces like l-dx.atlassian.net cannot be read by generic web readers. Use the confluence MCP server or the lpedia CLI, which own authentication, pagination, and URL building.

What are the limitations of Confluence full-text search?▼

CQL text ~ is stemmed, recall-oriented full-text search, so a single phrasing can miss relevant pages. Try several phrasings in both English and Japanese before concluding something is not documented.