Skip to content

3aKHP/prts-mcp

vts/v1.2.0 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 20d MCP Data & Storage
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

ai-agent anthropic arknights claude fanfiction game-data
+5 more
mcp mcp-server model-context-protocol prts prts-wiki

ReleasePort's take

Light signal
editorial:auto 13d

Release ts/v1.2.0 introduces get_event_summary(), get_story_summary(), and list_stories with include_summaries to provide LLM-generated narratives for events, stories, and summaries.

Why it matters: Test the new summary tools in development; they enable narrative overviews of event chapters, story chapters, and inline story listings via DeepSeek API integration.

Summary

AI summary

Added story chapter and event summary tools with LLM-generated narratives.

Changes in this release

Feature Medium

Adds get_event_summary() tool providing narrative overview of event chapters

Adds get_event_summary() tool providing narrative overview of event chapters

Source: llm_adapter@2026-05-21

Confidence: high

 —
Feature Medium

Adds get_story_summary() tool with three-tier fallback chain for summaries

Adds get_story_summary() tool with three-tier fallback chain for summaries

Source: llm_adapter@2026-05-21

Confidence: high

 —
Feature Medium

list_stories now accepts include_summaries parameter to display inline summaries

list_stories now accepts include_summaries parameter to display inline summaries

Source: llm_adapter@2026-05-21

Confidence: high

 —
Feature Medium

LLM summarization pipeline generates chapter and event summaries via DeepSeek API

LLM summarization pipeline generates chapter and event summaries via DeepSeek API

Source: llm_adapter@2026-05-21

Confidence: low

 —
Full changelog

Added

  • Story chapter summaries. Two new tools and one enhancement built on the
    previously-unused zh_CN/storyinfo.json index (1,945 entries), plus an
    LLM summarization pipeline in the data-source fork:
    • get_event_summary(event_id) — narrative overview of every chapter in an
      event, with chapter codes, tags, names, and summary text. When LLM event
      summaries are available (zh_CN/event_summaries.json), a full-dialogue
      V2 synopsis is prepended above the chapter listing.
    • get_story_summary(story_key) — single-chapter summary with a three-tier
      fallback chain: LLM long summary (zh_CN/summaries.json, 5~7:1 compression),
      official one-liner (zh_CN/storyinfo.json), and chapter storyInfo field.
    • list_stories now accepts include_summaries (bool). When true, each
      chapter line includes an indented summary below it.
  • LLM summarization pipeline (3aKHP/ArknightsStoryJson fork):
    scripts/summarize.py generates per-chapter (5~7:1) and per-event (10:1)
    summaries via DeepSeek V4 Flash API during CI release, injecting them into
    zh_CN.zip for transparent consumption by the MCP server.

Install

# Run directly (Streamable HTTP on :3000)
npx prts-mcp-ts

# Or install globally
npm install -g [email protected]

MCP endpoint: http://localhost:3000/mcp

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track 3aKHP/prts-mcp

Get notified when new releases ship.

Sign up free

About 3aKHP/prts-mcp

MCP Server for Arknights, querying the PRTS Wiki API and serving auto-synced operator archives and voice lines from game data. Designed for fan-creation (同人創作) AI agents. Python (stdio/Docker) and TypeScript (Streamable HTTP) implementations.

All releases →

Related context

Earlier breaking changes

  • vts/v1.0.0 Public tool surface frozen; 9 MCP tool names, parameters, and response formats locked.

Beta — feedback welcome: [email protected]