Expand model tag support: add GLM-5.1, simplify Anthropic IDs, scan tags anywhere in message

- Flink update_bars debouncing
- update_bars subscription idempotency bugfix
- Price decimal correction bugfix of previous commit
- Add GLM-5.1 model tag alongside renamed GLM-5
- Use short Anthropic model IDs (sonnet/haiku/opus) instead of full version strings
- Allow @tags anywhere in message content, not just at start
- Return hasOtherContent flag instead of trimmed rest string
- Only trigger greeting stream when tag has no other content
- Update workspace knowledge base references to platform/workspace and platform/shapes
- Hierarchical knowledge base catalog
- 151 Trading Strategies knowledge base articles
- Shapes knowledge base article
- MutateShapes tool instead of workspace patch

gateway/knowledge/trading/strategies/global-macro/fundamental-macro-momentum.md

@@ -0,0 +1,59 @@
+---
+description: "A systematic macro momentum strategy that buys assets favored by incoming macroeconomic trends and sells adversely affected assets, using four state variables to rank global equity indexes, currencies, and bonds."
+tags: [global-macro, momentum, systematic, multi-asset]
+---
+
+# Fundamental Macro Momentum
+
+**Section**: 19.2 | **Asset Class**: Global Macro | **Type**: Systematic momentum / Cross-sectional ranking
+
+## Overview
+
+This strategy aims to capture returns from the market's underreaction to changes in macroeconomic trends by buying assets favored by incoming macroeconomic trends and selling (shorting) assets adversely affected by them. It is a systematic (non-discretionary) strategy that can be applied across different asset classes — global equity indexes, currencies, government bonds, commodities, etc. — and exploits the tendency of prices to lag macroeconomic fundamentals.
+
+## Construction / Mechanics
+
+### State Variables
+
+Four macroeconomic state variables are used to characterize conditions in each country:
+
+1. **Business cycle trend**: estimated using 1-year changes in real GDP growth and CPI inflation forecast, each contributing with a 50% weight
+2. **International trade trend**: estimated using 1-year changes in spot FX rates against an export-weighted basket
+3. **Monetary policy trend**: estimated using 1-year changes in short-term interest rates
+4. **Risk sentiment trend**: estimated using 1-year equity market excess returns
+
+Note: different asset classes are affected by macroeconomic trends differently. For example, increasing growth is positive for equities and currencies but negative for bonds.
+
+### Portfolio Construction
+
+1. For a given asset class (e.g., global equity indexes), rank each country's asset using the values of the four state variables
+2. Construct a zero-cost long-short portfolio by going long the assets in the top decile and shorting those in the bottom decile
+3. Portfolios for various asset classes can be combined, e.g., with equal weights
+4. Holding period typically ranges from three to six months
+
+Multiple ranking and portfolio construction methods are available (see also Subsection 3.6 for multifactor portfolio construction approaches).
+
+## Return Profile / Objective
+
+Returns are driven by the market's systematic underreaction to macroeconomic trends: as trends gradually become recognized by market participants, prices adjust and the long positions appreciate while short positions decline. The strategy is fundamentally a momentum strategy but grounded in macroeconomic state variables rather than pure price momentum. Typical holding periods of 3–6 months mean lower turnover and transaction costs compared to short-term momentum.
+
+## Key Parameters / Signals
+
+- **Business cycle signal**: 1-year change in real GDP growth + CPI inflation forecast (50/50 weight)
+- **International trade signal**: 1-year change in spot FX rate vs. export-weighted basket
+- **Monetary policy signal**: 1-year change in short-term rates
+- **Risk sentiment signal**: 1-year equity market excess return
+- **Decile cutoffs**: top/bottom decile thresholds for portfolio construction
+- **Asset class weighting**: equal weight across combined asset class portfolios
+- **Holding period**: 3–6 months typical
+
+## Variations
+
+- **Single asset class**: apply the strategy only to equity indexes, only to currencies, or only to government bonds
+- **Broader factor set**: add additional state variables (e.g., current account balance, credit conditions)
+- **Discretionary overlay**: combine the systematic signal with analyst judgment
+- **CTA/managed futures extension**: apply the framework to futures contracts on equity indexes, bonds, and currencies
+
+## Notes
+
+Macro trading strategies constitute an investment style, not an asset class — they are not limited to any particular asset class or geographic region. The strategy can be classified as directional, long-short, or relative value depending on implementation. The three broad categories of macro strategies are discretionary macro, systematic macro, and CTA/managed futures. This strategy falls in the systematic macro category. Transaction costs are relatively modest given the 3–6 month holding period, but liquidity varies significantly across country assets. Care must be taken when applying this strategy to emerging markets where data quality and market access can be constraints.