Weapon: Give Your Agent a Real Skill Arsenal

The FROST V4.1 framework introduces a Weapon Registry that organizes agent skills into discoverable, composable, and versionable units. This registry enables agents to dynamically discover and compose skills, replacing hardcoded tool chains with declarative workflows. The system supports semantic versioning, metadata schemas, and composite weapons for exponential capability growth.

FROST Five-Dimensional Meta-Model Series · Part 1 Imagine you have an Agent. It's smart—it can write code, analyze data, search for information. But when you ask it: "Help me build a crawler to scrape e-commerce prices and analyze trends," it freezes. Not because it can't, but because it can't find what it knows. It has a bunch of skills scattered across the codebase, but no unified place telling it: "You know these things, use them as needed." This is the common problem with current Agent frameworks: skills exist, but they're unorganized. In FROST V4.1, we introduced the Weapon Registry . A Weapon is not a traditional "tool" or "function"—it's: A discoverable, composable, versionable skill unit. Think of it this way: python from core.armory import Armory Create a weapon registry armory = Armory Register a weapon armory.register name="web scraper", version="1.0.0", skill=WebScraperSkill , metadata={ "category": "data collection", "description": "Scrape web data", "input schema": {"url": "str", "selector": "str"}, "output schema": {"data": "list dict "}, } Register another one armory.register name="trend analyzer", version="1.0.0", skill=TrendAnalyzerSkill , metadata={ "category": "analysis", "description": "Analyze data trends", "input schema": {"data": "list dict "}, "output schema": {"trend": "str", "confidence": "float"}, } Every weapon has: web scraper , no conflicts with other skills 1.0.0 , supports upgrades and rollbacks js Find by category data skills = armory.find by category "data collection" = web scraper, api fetcher, file parser, ... Find by capability fuzzy match relevant = armory.find by capability "scrape web" = web scraper List all available weapons all weapons = armory.list all = web scraper, trend analyzer, code writer, ... Agents no longer need to hardcode "what I can do"—they dynamically discover which weapons they can use. Define a composite weapon armory.register composite name="ecommerce price tracker", components= "web scraper", "trend analyzer" , description="E-commerce price tracking and trend analysis", workflow=""" 1. Use web scraper to scrape price data 2. Pass data to trend analyzer for analysis 3. Return analysis report """ Use the composite weapon result = agent.execute "ecommerce price tracker", { "url": "https://example.com/product", "selector": ".price" } Composite weapons let Agent capabilities grow exponentially: | Dimension | Traditional Tool Calling | FROST Weapon Registry | |---|---|---| | Discovery | Hardcoded in source | Dynamic query, matched by need | | Versioning | None | Semantic versioning 1.0.0 → 1.1.0 | | Metadata | Simple function signatures | Full input/output schemas + categories + descriptions | | Composition | Manual glue code | Declarative composition, auto-generated workflows | | Testability | Need to mock entire tool chains | Each weapon tested independently, compositions also testable | Key difference: The Weapon Registry makes skills first-class citizens , not just appendages to code. Suppose you want to build an automated competitor analysis Agent : python Hardcode all steps class CompetitorAnalyzer: def init self : self.scraper = WebScraper Hard dependency self.analyzer = TrendAnalyzer Hard dependency self.reporter = ReportGenerator Hard dependency def analyze self, competitor url : data = self.scraper.scrape competitor url trend = self.analyzer.analyze data report = self.reporter.generate trend return report Problems: Declarative composition agent = Agent armory=armory Agent decides which weapons to use result = agent.execute "competitive analysis", { "competitor": "https://competitor.com", "metrics": "price", "features", "reviews" } Agent's internal logic automatic : 1. Query armory.find by capability "scrape web" → select web scraper 2. Query armory.find by capability "analyze trends" → select trend analyzer 3. Query armory.find by capability "generate report" → select report generator 4. Compose and execute, return result Advantages: This is FROST's design philosophy: Tools are passive; weapons are active. In FROST's family governance model: The Weapon Registry makes this process auditable, controllable, and evolvable . Weapons solve "what Agent can do." But how do Agents collaborate ? How do they know "task completed," "error occurred," "need help"? The answer is Event —Agent's nervous system. Next article preview: Event: Teaching Agents to "Shout Out" FROST: Giving agents lineage, memory, and honor.