Monitor Claude Code and Codex local usage logs for spend, model mix, and session burn with ccusage
Use ccusage when an agent operator needs to turn local Claude Code or Codex usage logs into spend and usage reports instead of manually reading raw JSONL files.
npx skills add agentskillexchange/skills --skill monitor-claude-code-and-codex-local-usage-logs-for-spend-model-mix-and-session-burn-with-ccusage
Tool: ccusage. This skill gives an agent a concrete reporting job: read local coding-agent usage logs and summarize cost hotspots, model mix, and session burn patterns into something a human can act on.
What this skill actually does
When to use it: invoke this when you need budget reviews, local usage audits, or per-session cost triage for Claude Code or Codex runs. Using this skill is different from using the product normally because the workflow is operational rather than exploratory: ingest the logs, generate the report, and surface the sessions or models that deserve follow-up.
Scope boundary: this is not a generic analytics dashboard listing and not a broad coding-agent platform entry. Its boundary is specific: analyze locally stored agent usage logs and turn them into cost and usage diagnostics with ccusage.