algebraic-effects

name	algebraic-effects
description	Implements algebraic effects and effect handlers for computational effects in typed languages.
version	1.0.0
tags	["effects","type-theory","semantics","handlers","monads"]
difficulty	advanced
languages	["ocaml","haskell","eff","python"]
dependencies	["effect-system","type-inference-engine"]

Algebraic Effects

Implements algebraic effects and effect handlers - a modular approach to computational effects where effects are described as operations and handled by handler functions.

When to Use This Skill

• Building languages with composable effects
• Implementing async/await, state, exceptions, logging, nondeterminism
• Researching effect systems and handlers
• Creating domain-specific effect languages

What This Skill Does

1. Effect Signature Definition: Defines effects as sets of operations with types
2. Effect Handler Implementation: Implements handlers that interpret effect operations
3. Deep vs Shallow Handlers: Supports both continuation-based deep handlers and single-shot shallow handlers
4. Effect Inference: Infers effect types from code using effect inference algorithms

Key Concepts

Concept	Description
Effect Operation	Abstract description of an effect (e.g., State.get, State.put)
Effect Handler	Function that interprets effect operations
Resumption	Continuation passed to handler for resuming computation
Effect Row	Collection of effects tracked in type system

Tips

• Start with simple effects (state, exceptions)
• Use deep handlers for full resumption
• Consider effect inference to reduce annotation burden
• Handlers compose via delegation

Related Skills

• effect-system - Effect type tracking
• monad-transformer - Alternative effect composition
• type-inference-engine - Effect inference

Canonical References

Reference	Why It Matters
Plotkin & Pretnar "Handlers of Algebraic Effects" (2009)	Foundational effect handlers paper
Bauer & Pretnar "Programming with Algebraic Effects and Handlers"	Eff language implementation
Leijen "Koka: Programming with Row Polymorphic Effect Types" (2014)	Effect rows in practice
Kammar et al. "Handlers in Action" (ICFP 2013)	Comprehensive handler study

Tradeoffs and Limitations

Approach	Pros	Cons
Deep handlers	Full resumption, composable	Complex continuation management
Shallow handlers	Simpler implementation	Limited expressiveness
Monadic encoding	Familiar, pure	Effets verbosity, lack of local reasoning

Assessment Criteria

Criterion	What to Look For
Type safety	Effect operations well-typed
Handler correctness	Operations correctly interpreted
Performance	Minimal overhead vs monadic approach
Composability	Handlers can be layered

Quality Indicators

✅ Good: Type-safe, correct handlers, good performance, composable ⚠️ Warning: Partial type safety, some handler issues ❌ Bad: Type errors, incorrect handler behavior

Research Tools & Artifacts

Algebraic effect implementations:

Tool	Language	What to Learn
Eff	OCaml	Original implementation
Koka	Koka	Effect types
Frank	Frank	Handler calculus
Idris 2	Idris	Effects

Research Frontiers

1. Effect Inference

• Goal: Reduce annotation burden
• Approach: Row polymorphism, effect variables
• Papers: "A Theory of Effect Inference" (Lucassen & Gifford)

2. Algebraic Effects for Concurrency

• Goal: Composable concurrency primitives
• Approach: Effect handlers for async/await
• Papers: "Asynchronous Effect Handlers" (2019)

Implementation Pitfalls

Pitfall	Real Consequence	Solution
Continuation leak	Memory issues	Proper resumption handling
Effect row ambiguity	Unclear effect set	Explicit row variables
Handler ordering	Wrong semantics	Define evaluation order