QFM122: Elixir Reading List - July 2026

phxagents is a specialized Claude Code plugin that enforces Phoenix and Elixir best practices through 26 "Iron Laws" (production safety rules) and a four-phase workflow (plan → work → review → compound) to prevent common AI-generated code mistakes like unoptimized Repo queries, bare rescues, and missing Ecto transactions. The plugin integrates runtime introspection via Tidewave MCP, parallel code review by four specialist agents, and supply-chain security auditing for Hex dependencies, adapting to your specific Phoenix version and project tooling rather than applying generic guidance.

The author implements automated Elixir quality checks to prevent AI-generated code from gradually degrading project standards through repeated low-quality patterns. The approach combines compiler checks (warnings-as-errors, format validation, dependency auditing), static analysis tools like Credo with custom rule sets, and test quality linters to mechanically enforce coding standards while still requiring human review of AI-generated changes.

Tincture is a pure Elixir library for PDF generation that implements typographic quality rendering entirely in Elixir/OTP without external dependencies, featuring TeX hyphenation, Knuth-Plass line breaking, TrueType/OpenType font embedding with subsetting, and support for PDF/UA and PDF/A standards. Built as a modernized fork of erlguten with preserved attribution, it includes advanced capabilities like GPOS kerning, GSUB ligatures, AES-256 encryption, forms, and digital signatures, with conformance verified against PDF/UA-1 and PDF/A-2 standards.

Large language models achieve their highest performance on Elixir compared to 19 other languages tested on AutoCodeBench, with a 97.5% upper bound and individual models scoring ~30 points higher on Elixir than their overall average, despite Elixir having roughly 10 times less training data than Python. This superior performance persists across different model architectures, reasoning modes, and even base models without instruction tuning, suggesting the advantage stems from Elixir's language design—particularly its immutability and explicit data flow—which aligns with how language models process code rather than from memorization of training examples.

Credence is a semantic linter for Elixir that identifies and automatically fixes code that compiles and passes tests but is written inefficiently or idiomatically incorrect, such as patterns copied from Python or JavaScript. It operates in three sequential rounds—Syntax (fixes parsing errors), Semantic (resolves compiler warnings), and Pattern (applies ~117 idiomatic rules to rewrite awkward code like Enum.sort |> Enum.reverse to Enum.sort(:desc))—with a strict guarantee that every pattern rule produces a safe, correct fix rather than just warnings.