midea-ai/semaclaw/wiki

Claude Code and agent skill by midea-ai/semaclaw

Trust tier 2
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Save learning/research documents to the personal wiki knowledge base and search existing content

Install

npx vskill@latest install midea-ai/semaclaw/wiki

SKILL.md

Wiki Knowledge Base Management

The user's personal knowledge base is maintained with the semaclaw wiki command. The wiki directory is located at {home}/semaclaw/wiki/, where {home} is the user's home directory (~ on macOS/Linux, %USERPROFILE% on Windows). The knowledge base is organized by topic folders, with each document as a Markdown file.

Writing New Documents to Wiki

Full Workflow
  1. View the directory structure to understand existing topic categories:

`` semaclaw wiki tree ``

  1. Determine the category:
  • Content belongs to an existing directory → save directly there
  • No suitable directory exists → create a new topic directory under the closest parent:

`` semaclaw wiki mkdir "new/directory/path" ``

  • Completely uncertain → stage in inbox/ and inform the user to categorize later
  1. Save the document (pipe full Markdown content via heredoc to stdin):

```bash cat <<'WIKI_EOF' | semaclaw wiki save --path "directory/filename.md" --tags "tag1,tag2" # Document Title

Document content... WIKI_EOF ```

Example
bash
# Save an article about Rust async
cat <<'WIKI_EOF' | semaclaw wiki save --path "programming/rust/async-runtime.md" --tags "rust,async,tokio"
# Rust Async Runtime Explained
## Core Concepts
Tokio is the most popular async runtime for Rust...
WIKI_EOF

Output JSON: {"path": "programming/rust/async-runtime.md", "action": "created"}

Organizing Existing Documents

Use this workflow when the user wants to organize, classify, or tidy up documents already on disk into the wiki.

Critical rules:

  • Always use `cp` to copy files — never rewrite or regenerate document content. Rewriting wastes time and risks altering the original.
  • Use mv only when the user explicitly asks to move (e.g. "move it", "don't keep the original").
  • Never rewrite document body content during organization. Only touch the YAML frontmatter.
Full Workflow
  1. Read the document title and first ~300 characters to infer the topic — do not read the entire file.
  1. View the wiki directory structure:

`` semaclaw wiki tree ``

  1. Determine the target category using the same rules as saving:
  • Matches an existing directory → use it
  • No match → create with semaclaw wiki mkdir "path"
  • Uncertain → place in inbox/
  1. Copy the file to the wiki (do not rewrite content):

``bash cp "/source/path/document.md" "{home}/semaclaw/wiki/category/filename.md" # or on Windows: copy "source" "dest" ``

  1. Edit the YAML frontmatter of the copied file to add/update tags and metadata. If no frontmatter exists, prepend it:

```markdown


tags: [tag1, tag2] source: /original/path/document.md


``` Use the Edit tool to make this change — do not regenerate the file.

When to use mv instead of cp

Only when the user explicitly says "move", "don't keep the original", or similar. Otherwise always default to cp.

Searching Existing Knowledge

bash
# Search by title/filename
semaclaw wiki search "rust async"
# Filter by tags
semaclaw wiki search "" --tags "tokio"

When to Trigger

Write workflow — user says:

  • "add to wiki", "save to knowledge base", "archive this"
  • "save this to wiki", "note it in the knowledge base"

Organize workflow — user says:

  • "put this xxx file in wiki"
  • "file these into the knowledge base"
  • "sort these files into the wiki"

Search workflow — user says:

  • "check my notes on X", "do I have anything in my wiki about X"
  • "look up X in my knowledge base", "search my wiki for X"
  • "what did I save about X", "have I documented X before"
  • "find my notes on X", "pull up what I know about X"

Filename Conventions

  • Concise description of the topic (async-runtime.md not notes-on-async-runtime-learning.md)
  • No more than 40 characters

Security

  • Security scanPassed, 93 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
64
Repository forks
11
7-day trend score
21

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/midea-ai/semaclaw/wiki/badge)](https://verified-skill.com/skills/midea-ai/semaclaw/wiki)

Something wrong with this skill or its page? Tell us.