tokencast CLAUDE.md

tokencast CLAUDE.md is an instructions file for coding agents from krulewis/tokencast. It costs 1,446 tokens per session, scanned A, original, MIT.

A token-cost estimator for Claude Code plans that learns from actual usage to improve its estimates. Claude Code is a programming assistant, and tokens are the text units used to calculate model usage.

In plain words
What is it for?
Use it when planning development work, estimating Anthropic API costs, reviewing token budgets, or analyzing recorded usage; the input identifies it mainly as an MCP server and reference documentation.
Why use it?
It helps compare expected API costs with actual costs and adjusts future estimates using calibration data.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/krulewis/tokencast/claude-md
Clone the repo
git clone --depth 1 https://github.com/krulewis/tokencast

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for tokencast CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/krulewis/tokencast/claude-md.svg)](https://agentmods.dev/instructions/krulewis/tokencast/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/krulewis/tokencast/claude-md"><img src="https://agentmods.dev/badge/instructions/krulewis/tokencast/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,446 This file is loaded in full into every session.
When invoked 1,446 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.01446 $0.01446
Opus 5 $0.00723 $0.00723
Sonnet 5 $0.00289 $0.00289
Haiku 4.5 $0.00145 $0.00145

Measured 4d ago against content hash decb5420dbec, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

tokencast CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 85 lines

How it starts

The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.

tokencast

A Claude Code skill that automatically estimates Anthropic API token costs when a development plan is created, and learns from actual usage over time to improve accuracy via calibration factors.

Repo

  • GitHub: krulewis/tokencast
  • PyPI package version: 0.1.4
  • Distribution: MCP server via uvx tokencast (registered in MCP Registry as io.github.krulewis/tokencast)
  • SKILL.md: Retained as algorithm reference doc, but no longer installed as a Claude Code skill. MCP server is the primary interface.

Key Files

Path Purpose
SKILL.md Algorithm reference doc — estimation formulas, output template (v2.1.0, no longer installed as skill)
references/heuristics.md Token budgets, pipeline step decompositions, complexity multipliers, parallel discount parameters — all tunable parameters live here
references/pricing.md Model pricing per million tokens, cache rates, step→model mapping
references/calibration-algorithm.md Calibration algorithm documentation
references/examples.md Worked estimation examples
scripts/tokencast-learn.sh Stop hook — reads session JSONL, computes actuals, calls update-factors.py
scripts/tokencast-midcheck.sh PreToolUse hook — warns if trending toward pessimistic band
scripts/tokencast-agent-hook.sh PreToolUse+PostToolUse hook — writes sidecar timeline for per-agent step attribution
scripts/update-factors.py Computes and persists calibration factors from session data
scripts/sum-session-tokens.py Parses session JSONL, computes per-step actuals via sidecar
scripts/calibration_store.py Storage abstraction for history.jsonl and factors.json
scripts/tokencast-status.py Calibration health dashboard (v2.0)
calibration/ Calibration data — gitignored; contains history.jsonl, factors.json, active-estimate.json
src/tokencast/__init__.py Package exports (eager imports — lazy migration was not needed; CI fix was GNU xargs compat)
src/tokencast/api.py Public API layer — dict-based routing; estimate_cost(), report_session(), report_step_cost()
src/tokencast/estimation_engine.py Core estimation: compute_estimate(params, calibration_dir)
src/tokencast/file_measurement.py File size bracket measurement: measure_files(), assign_bracket()
src/tokencast/pricing.py Cost computation: compute_cost_from_usage(usage, model)
src/tokencast/heuristics.py Tunable parameters (derived from references/heuristics.md)
src/tokencast/session_recorder.py build_history_record() — shared by shell and MCP paths
src/tokencast/step_names.py Step name resolution: resolve_step_name() (handles alias mapping)
src/tokencast/telemetry.py Opt-out PostHog telemetry (endpoint hardcoded, install ID at ~/.tokencast/install_id, no-telemetry file at ~/.tokencast/no-telemetry)
src/tokencast/calibration_store.py Storage abstraction for history and factors
src/tokencast/parse_last_estimate.py Reconstitution of minimal estimates from last-estimate.md (package module)
src/tokencast/tokencast_status.py Calibration health dashboard utilities (package module)
src/tokencast/update_factors.py Calibration factor computation and persistence (package module)
src/tokencast_mcp/server.py MCP server: main(), build_server(), tool dispatcher
src/tokencast_mcp/tools/ MCP handlers: estimate_cost, get_calibration_status, get_cost_history, report_step_cost, report_session, disable_telemetry
src/tokencast_mcp/tools/disable_telemetry.py Handler for disable_telemetry tool — creates ~/.tokencast/no-telemetry file (v0.1.5+)
docs/wiki/ GitHub wiki source — Home, How-It-Works, Installation, Configuration, Calibration, Roadmap, Attribution
docs/attribution-protocol.md Framework-agnostic attribution protocol spec (v1)
docs/phase-1-execution-plan.md Owner decisions and story inventory (Phase 1 completed)
README.md Repo root README (PyPI package docs)
pyproject.toml Package metadata; entry point tokencast-mcp = "tokencast_mcp.server:main"

Read the full file on GitHub · 85 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 85 lines · 1,446 tokens per session scan A decb5420dbec

Subscribe to this mod's changes

tokencast CLAUDE.md is an instructions file published in the GitHub repository krulewis/tokencast (2 stars, last pushed 4mo ago), licensed MIT. It adds 1,446 tokens to every session, about $0.0072 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens