claude-credit-auditlisted

# Claude Credit Audit Reads a repo's CI config, scripts, and the user's subscription plan to **forecast, ahead of time**, which agent calls would burn Agent SDK credit *if/when* Anthropic's announced 2026-06-15 billing change takes effect, whether the monthly credit limit would run out (and when), and how to cut cost. Unlike after-the-fact usage trackers, it scans workflows/scripts to predict. > **Status (2026-06-16): the billing change is paused** ([Anthropic note](https://support.claude.com/en/articles/15036540-use-the-claude-agent-sdk-with-your-claude-plan) — "for now, nothing has changed"). When presenting results, make clear this is a forecast of what would happen if/when the change resumes, not current billing. ## Locale Default to **English** (`--lang en`). Use **Chinese** (`--lang zh`) when the user's request is primarily Chinese, the target audience is Chinese, or the user asks for a Chinese report. Present your spoken summary in the same language you pass to `--lang`. ## When to trigger The user wants the credit cost of a repo's Claude/agent calls after the 6/15 change, or wants to audit `claude -p`, `claude-code-action`, Agent SDK, or Anthropic SDK calls in CI / workflows / scripts. ## How to run (you execute it — do not make the user type commands) This skill bundles a Python CLI. **Prerequisite:** Python ≥ 3.10 with PyYAML on PATH; the wrappers check and tell the user to `pip install pyyaml` if it's missing. Run the wrapper from **this skill's directory** (i