turntax

turntax

Coding agents re-send their instruction files with every turn. turntax measures that fixed per-turn cost for a repository, for every agent that reads it, and shows where to cut it.

npx turntax

Try it in the browser, with nothing uploaded: https://torutdev.github.io/turntax/

Supported agents

Each agent's documented loading rules decide what is sent on every turn, what loads only when the agent works in a folder or touches a matching file, and what is ignored.

Agent	Files it reads from the repository
Aider	files listed under read: in .aider.conf.yml (CONVENTIONS.md only when listed or passed with --read)
Amp	AGENTS.md (or AGENT.md / CLAUDE.md), nested AGENTS.md, @-mentioned files
Claude Code	CLAUDE.md, CLAUDE.local.md, .claude/rules, skills, commands, subagents, AGENTS.md when there is no CLAUDE.md
Cline	.clinerules/, .cline/rules/
Continue	.continue/rules/
Cursor	.cursor/rules/*.mdc, .cursorrules, AGENTS.md, CLAUDE.md
Gemini CLI	GEMINI.md
GitHub Copilot	.github/copilot-instructions.md, .github/instructions/, AGENTS.md
Goose	.goosehints, AGENTS.md
JetBrains Junie	.junie/AGENTS.md, or AGENTS.md with .junie/rules/, or .junie/guidelines
Kiro	.kiro/steering/, AGENTS.md
OpenAI Codex	AGENTS.md, AGENTS.override.md, .agents/skills
OpenCode	AGENTS.md, or CLAUDE.md when there is none
Roo Code	.roo/rules/, .roo/rules-{mode}/, .roorules, AGENTS.md
Windsurf (Devin Desktop)	.devin/rules/, .windsurf/rules/, .windsurfrules, AGENTS.md
Zed	the first of .rules, .cursorrules, .windsurfrules, .clinerules, .github/copilot-instructions.md, AGENT.md, AGENTS.md, CLAUDE.md, GEMINI.md

Agents that load exactly the same files share one row in the report:

always  o200k/turn  on-demand  model                tokens/turn  counted by                   100 turns  cached   agents
4       272         53         (no selected model)                                                                Claude Code
1       63          0          GPT-6.1 Sol          63           estimate (o200k_base proxy)  $0.01      $0.0006  OpenAI Codex, Roo Code, Kiro, JetBrains Junie, Goose, OpenCode
1       17          0          (no selected model)                                                                Gemini CLI
4       297         37         GPT-6.1 Sol          297          estimate (o200k_base proxy)  $0.06      $0.0030  Cursor
2       74          8          GPT-6.1 Sol          74           estimate (o200k_base proxy)  $0.01      $0.0007  GitHub Copilot
2       75          0          GPT-6.1 Sol          75           estimate (o200k_base proxy)  $0.01      $0.0008  Windsurf (Devin Desktop)
1       11          0          GPT-6.1 Sol          11           estimate (o200k_base proxy)  $0.0022    $0.0001  Zed
1       63          19         GPT-6.1 Sol          63           estimate (o200k_base proxy)  $0.01      $0.0006  Amp

Advice (savings are estimates in o200k tokens per turn)
  HIGH duplicate CLAUDE.md, AGENTS.md [saves ~48 tokens/turn]
    3 identical lines are loaded twice every turn by cursor. Keep them in one file.
  HIGH contradiction AGENTS.md, CLAUDE.md [saves ~11 tokens/turn]
    Conflicting package manager: "Always use npm ..." (AGENTS.md:3) vs "Always use pnpm ..." (CLAUDE.md:5).
  MEDIUM scopable-section CLAUDE.md:11 [saves ~128 tokens/turn]
    Section "API package" is mostly about packages/api/. Move it to packages/api/AGENTS.md ...

(Excerpt of the output for test/fixtures/mixed with --models gpt-6.1-sol.)

What it counts, and how

Model family	How tokens are counted	Label
OpenAI GPT-5.x, gpt-oss	Offline with o200k_base (the encoding OpenAI's tiktoken maps them to)	exact
OpenAI GPT-6.x	Offline with o200k_base; OpenAI has not published a newer encoding	estimate
Claude	Only Anthropic's count_tokens API, with your key and --claude-api	API count
Gemini	Only Google's countTokens API, with your key and --gemini-api	API count

Claude's tokenizer is not public, and Anthropic's docs note that newer Claude models produce noticeably more tokens than older ones for the same text. turntax never shows an OpenAI count as a Claude or Gemini number: without a key those rows say "needs API key". The o200k column is a labelled reference count for every file.

Cost per N turns = tokens per turn × N × input price. The "cached" column assumes every turn reads the context from the prompt cache. Prices come from a dated table with a source link per model (turntax --list-models); pass --prices my-prices.json to use your own.

See docs/how-it-counts.md for the loading rules per agent and their sources.

Advice

All checks are deterministic. Each one shows an estimated saving in o200k tokens per turn.

• duplicate: identical lines in two files. A per-turn saving only when one tool loads both; otherwise flagged as drift risk.
• contradiction: conflicting rules reaching the same agent (package manager, indentation, quotes, semicolons).
• scopable-section: a section mostly about one directory that could load only there (nested file, path-scoped rule, applyTo).
• oversized-section: a large always-loaded section that could become an on-demand skill.
• size-limit: documented limits (CLAUDE.md under 200 lines, Codex's 32 KiB cut-off, Windsurf's 12,000 characters, Cursor's 500 lines).
• misconfigured: files a tool ignores, such as .md files in .cursor/rules or Windsurf rules without a trigger.
• stale-reference: referenced files or npm run scripts that no longer exist.
• non-english: non-English sections. The token multiplier is measured only against an English version you supply (--english CLAUDE.md=CLAUDE.en.md) or, if you opt in, a Claude translation (--translate).

Options

turntax [path]
  --tools <ids>             agents to report (default: every agent that reads a file); see --help
  --models <ids>            see --list-models (default: claude-sonnet-5.5, gpt-6.1-sol, gemini-3.8-flash)
  --turns <n>               cost projection (default 100)
  --json | --md | --badge   output format (default: terminal table)
  --out <file>              write the output to a file
  --max-tokens <n>          exit 1 when any tool loads more than n o200k tokens per turn (CI gate)
  --claude-api              exact Claude counts (ANTHROPIC_API_KEY)
  --gemini-api              Gemini counts (GEMINI_API_KEY)
  --english <file=en>       measure a language multiplier against your English version
  --translate               opt-in Claude translation for that measurement (needs --claude-api)
  --prices <file.json>      your own price table
  --section-tokens <n>      threshold for oversized sections (default 600)

Privacy

Offline by default: no network calls, no telemetry. File contents are sent only when you pass --claude-api, --gemini-api or --translate, and only to that vendor's API. Keys are read from environment variables and sent only in request headers.

Web page

web/ is a static page that runs the same engine in the browser with the offline tokenizer only. Drop a repository folder or paste files; nothing is uploaded.

npm run build:web

Badge

npx turntax --badge

The badge shows the largest always-loaded total among the tools found, counted with o200k_base.

Limits

• Assumes the agent starts at the repository root. Files outside the repository (user-level ~/.claude/CLAUDE.md, global rules, team rules in a dashboard) are not counted.
• Counts instruction files only, not the agent's own system prompt, tool definitions or MCP servers.
• Loading rules change. Each rule cites the vendor page and the date it was checked; please open an issue when one is out of date.

Contributing

Loading rules and prices go out of date. Corrections with a link to the vendor's page are the most useful contribution; see CONTRIBUTING.md.

License

MIT