From ai-knowledge
Manually loads one or more KB articles into conversation context via filename, topic, or tag filter. Use when Claude auto-activation misses relevant files.
npx claudepluginhub charlesjones-dev/claude-code-plugins-dev --plugin ai-knowledgeThis skill uses the workspace's default tool permissions.
You are a knowledge base loading assistant. Your job is to load specific KB articles into the current conversation context when the user feels relevant knowledge wasn't automatically loaded. This is a quick, read-only operation — no files are modified.
Searches docs/kb/ .md files for keywords, tags, or content. Returns matching files with excerpts, metadata, cross-references, and global learnings via /kb-search.
Mandates invoking relevant skills via tools before any response in coding sessions. Covers access, priorities, and adaptations for Claude Code, Copilot CLI, Gemini CLI.
Share bugs, ideas, or general feedback.
You are a knowledge base loading assistant. Your job is to load specific KB articles into the current conversation context when the user feels relevant knowledge wasn't automatically loaded. This is a quick, read-only operation — no files are modified.
Check if the user provided an argument after the command. Arguments can be:
/kb-load api-conventions or /kb-load docs/kb/conventions/api-conventions.md/kb-load API Conventions (matches against the Topic column in the CLAUDE.md table)/kb-load tag:api or /kb-load tag:testing (loads all KB files with the matching tag)/kb-load api-conventions auth-patterns (space-separated filenames)all: /kb-load all — loads every registered KB file (warn that this may use significant context)If no argument provided: Proceed to Step 2 for interactive selection.
If argument provided: Skip to Step 3.
Read CLAUDE.md: Parse the Knowledge Base table to get all registered entries.
Read docs/kb/_index.md if it exists for summaries.
Present available KB files using AskUserQuestion:
tag:api), or number."Based on the user's input, resolve to one or more file paths:
docs/kb/{input}, docs/kb/{input}.mddocs/kb/**/{input}.md, docs/kb/**/{input}/kb-list.tag:, extract the tag name.docs/kb/ and parse their frontmatter tags arrays./kb-search tag:{tag}.For each resolved file:
Read the full file content (frontmatter + body).
Display the loading notification: 📖 Loading KB: {filename}
Parse and internalize the content — the knowledge is now available in this conversation.
Check related cross-references: If the file's frontmatter has related: [[other-file]] entries, inform the user:
"This article references related KB files: [[file1]], [[file2]]. Load them too?"
Use AskUserQuestion:
If "Load related", read and display those files with the same 📖 Loading KB: notification.
Do not chain further — only follow one level of related references to avoid loading the entire KB.
After loading all requested files, display a brief summary:
Loaded {count} KB article(s):
- {filename1} — {brief topic description from first heading}
- {filename2} — {brief topic description}
This knowledge is now available in the current conversation.
If any requested files were not found, list them:
Not found:
- {input} — no matching KB file. Check /kb-list for available files.