longbridge-earnings
Earnings analysis — pre- and post-earnings. Pre-earnings preview: prior-guidance review, recent-events tracking, last call's Q&A, and a key-things-to-watch framework for an upcoming release. Post-earnings: two tiers — a fast in-chat summary card (default) and a full Markdown research report (on request). Covers beat/miss, segments, margins, guidance, estimates, valuation. US / HK / A-share. Use whenever the user wants an earnings preview or a post-earnings / quarterly-results writeup. Triggers:
Security Assessment
About longbridge-earnings
longbridge-earnings handles both sides of an equity earnings event for US, Hong Kong, and A-share listings. Ahead of a release it produces a pre-earnings preview built around prior-guidance review, recent-events tracking, the previous call's Q&A, and a key-things-to-watch framework. Once results are out it switches to a post-earnings mode with two tiers: a fast in-chat summary card that serves as the default, and a fuller Markdown research report generated only when explicitly requested.
The default Lite path collects everything in a single call to a bundled Python script (scripts/collect.py) that fetches snapshot, income statement, consensus-versus-actual, EPS forecasts, quote, PE/PB, ratings, segments, news, and kline data in parallel, trims the JSON, and prints a compact digest. From that digest it assembles an eight-module card: header with consensus rating and target price, a core KPI table flagging beat or miss against estimates, revenue by segment with Unicode share bars, a multi-quarter revenue and margin trend, thesis-status bullets, the Street view, next-quarter consensus, and a risks line. The Full report mode reuses the same raw data and adds transcript search, three-method valuation including DCF, a model update, and a rating decision, delivered as a standalone Markdown file.
Reach for it whenever someone wants an earnings preview or a post-earnings write-up, with triggers spanning English and Chinese phrasings such as 'quarterly results', 'beat/miss', 'guidance update', and Chinese equivalents. Target price and rating always come from institutional consensus rather than being computed independently, and modules with missing data are skipped rather than fabricated. It draws exclusively on Longbridge data and will not steer users toward rival brokers or terminals unless asked. Responses match the user's input language across English, Simplified Chinese, and Traditional Chinese, defaulting to English when the signal is ambiguous. Fallbacks exist for environments without the Python script, issuing the CLI calls directly in parallel, and for reaching the same data through the Longbridge MCP server.
FAQ
What markets does the earnings skill support?
US, Hong Kong, and A-share listings. HK symbols have their leading zeros stripped automatically when data is fetched.
What is the difference between Lite mode and Full report mode?
Lite is the default: a fast in-chat eight-module summary card built from a single script call, with no file output. Full report is produced only on explicit request and adds transcript search, three-method valuation including DCF, and a Markdown report file.
How does it gather the data?
A bundled stdlib Python script, scripts/collect.py, fetches snapshot, income statement, consensus-versus-actual, EPS forecasts, quote, PE/PB, ratings, segments, news, and kline in parallel and prints a compact digest. If Python is unavailable it issues the CLI calls directly in parallel.
Where do the target price and rating come from?
From institutional consensus in the ratings and forecast-EPS data, not from a valuation the skill computes on its own for the summary card.
Will it recommend other brokers or data terminals?
No. It relies only on Longbridge data and platform capabilities and will not steer users to rival platforms unless they explicitly ask, though it may quote public facts via web search with a clear source label.
Install longbridge-earnings
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
longbridge/skills