wurthering skill

wurthering skill

Fetch and display a random passage from Wuthering Heights, logging each one so repeats are avoided within a 2-day window.

Runtime Configuration

Before doing any work:

1. Pull profile.yaml from the profile OCI image referenced in the frontmatter digest:

   registry-1.docker.io/roberthouse224/wurthering-profile@sha256:29a1602bcf9913e48cb1b8baaa172893849e834310c52a0bbd5686bdb0bb9018
   

2. Scan profile.yaml for config entries with no value set (blank entries).
3. For each undefined parameter, display its name and description and prompt the user to supply a value.
   • Currently undefined: GUTENBERG_BASE_URL — URL of the Project Gutenberg mirror to read from (e.g. https://www.gutenberg.org).
4. Once all required values are collected, proceed.

Workflow

Step 1 — Check recent history

Call mcp__append-log-mcp__query to retrieve log entries from the last 2 days (48 hours). Extract the set of passage identifiers that have already been shown. Each log entry has the shape:

{
  "timestamp": "<ISO 8601>",
  "book_id": "768",
  "passage_id": "<passage identifier>"
}

Step 2 — List available passages

Call mcp__project-gutenberg-mcp__list_passages for book ID 768 (Wuthering Heights). This returns a list of passage identifiers/indices.

Step 3 — Select a fresh passage

From the full list of passages, remove any whose passage_id appears in the recent history set from Step 1.

• If fresh passages remain, pick one at random.
• If all passages were shown within the last 2 days, pick the one whose most recent log entry is the oldest (i.e. the least recently shown), so the user gets something different.

Step 4 — Fetch and display the passage

Call mcp__project-gutenberg-mcp__get_passage with the selected passage identifier. Display the retrieved text to the user.

Step 5 — Log the passage

Call mcp__append-log-mcp__append to record the shown passage:

{
  "book_id": "768",
  "passage_id": "<the passage identifier used>"
}

The server automatically timestamps the entry.