The Elements of Agent Style

Make your AI agent write like a tech pro.

21 rules: 12 from classic writing guides, 9 from field observation of LLM output, 2022-2026.

Before / after  ·  What it is  ·  Does it work?  ·  Use

Before and After

The top row is the identity: 21 rules, 12 canonical plus 9 field-observed; bordered red marks the two critical rules (RULE-01 curse-of-knowledge and RULE-H citation discipline). The bottom three rows are the mechanism, each pulled from a real v0.3.0 bench draft pair (same prompt, independent generations with and without the ruleset): a product description on Gemini 3 Flash (8 → 0 violations), a design-doc section on Claude Opus 4.7 (14 → 7), and a paper related-work section on Gemini 3 Flash (6 → 4; the source drafts are anchored to three prompt-named benchmarks (AgentBench, BFCL, tau-bench) so the style delta measured by agent-style review is isolated from fabricated-citation noise, though the displayed baseline snippet stops before the third name to keep the panel to three sentences). Each row shows the AI draft on the left with rule-violating phrases highlighted, the specific rules that fire in the middle, and the revised version on the right at the same length budget, same register, same document type.

What It Is

A curated set of English writing rules formatted for AI coding and writing agents to follow at generation time, not as a post-hoc linter.

Scope

Two Rule Groups, Peer Inputs

The 21 rules split by origin. The agent reads both as equal peers; no group is annotated as higher-priority than the other.

• RULE-01..12 (canonical). Distilled from Strunk & White, Orwell, Pinker, and Gopen & Swan. Each rule cites its source by chapter, section, or essay rule; I verified every citation against the original works. Full names in the table below.
• RULE-A..I (field-observed). Patterns I logged from AI output across dozens of writing projects (papers, grant proposals, technical documentation, agent configs) and code releases, 2022 to 2026. Transparent attribution: these are not literature-backed but capture LLM-specific failure modes that the canonical set alone misses (RULE-H citation discipline is the critical one). Full names in the table below.

Named in homage to Strunk & White's The Elements of Style (1918/1959), one of the four canonical sources.

Canonical Rules (RULE-01 through RULE-12)

Sourced from writing authorities; each rule cites its source by chapter, section, or essay rule.

Field-Observed Rules (RULE-A through RULE-I)

Structural patterns logged from LLM output across research, proposal, documentation, and agent-configuration work, 2022 to 2026. Treated as peer input to the canonical rules when an agent consumes the ruleset.

Escape hatch (Orwell 1946 Rule 6): "Break any of these rules sooner than say anything outright barbarous." Rules are guides to clarity, not ends in themselves.

See RULES.md for the full per-rule blocks with BAD/GOOD examples, enforcement tier, and rationale.

Does It Work?

Sanity bench on 10 fixed prose tasks (2 PR descriptions, 1 design-doc section, 1 commit message, 4 paper sections, 1 product description, 1 NSF-style specific aim), 2 generations per condition, flagship models drafting with agent-style loaded at generation time vs not. Three models shown: Claude Opus 4.7 dropped 45% (105 → 58 violations); OpenAI GPT-5.4 via Codex CLI dropped 45% (51 → 28); Gemini 3 Flash dropped 82% (79 → 14). The GitHub Copilot CLI runner was tested but excluded from the figure because its -p programmatic mode does not load instruction files (+3% noise; see CHANGELOG.md Notes). See docs/bench-0.3.0.md for the full scorecard across all 4 runners plus per-runner drafts under docs/bench-0.3.0-drafts/, and RELEASING.md "Bench (Local Only)" for reproduction steps.

Numbers are directional, not statsig: 10 tasks × 2 generations × 2 conditions = 40 calls per runner. The takeaway is that the ruleset reduces mechanical AI-tell density across every frontier family where the instruction file actually reaches the model's context; the exact size of the drop varies with each model's baseline (heavier in models that emit more long sentences / em-dashes by default, smaller or zero in already-clean models).

Who It Is For

• AI coding and writing agents: Claude Code, Codex, GitHub Copilot, Cursor, Aider, Anthropic Skills, Kiro, Replit Agent, Windsurf, Amazon Q Developer, JetBrains AI Assistant, Ollama, Continue.dev, Tabnine, OpenCode, OpenAI Agents SDK skills, and any AGENTS.md-compliant tool. Each agent reads the ruleset as part of its system prompt or project config.
• Maintainers of those agents' configurations who want generated prose to follow literature-backed writing conventions rather than reproduce common LLM tell-patterns.

Use

Two ways to enforce the rules in your AI agent's output. You typically use both:

• 1. Soft enforcement (default): the rules are loaded at generation time, so the agent tries to follow them while writing the first draft. One enable command per tool; then the agent reads agent-style's 21 rules as part of its context.
• 2. Skill (opt-in second pass): after the draft exists, you ask for a review. style-review audits the draft against the same 21 rules and, on your confirmation, writes a polished copy beside the original. Your source file is never touched.

Install the CLI once; then pick the paths you want.

pip install agent-style                              # Python users
# or: npm install -g agent-style                     # Node users
# or: npx --yes agent-style@0.3.6 <subcommand>       # no install needed

1. Soft Enforcement — rules at generation time