Budi analytics

Skill scoobydont-666/shared-claude-skills/skills/budi-analytics

budi token/cost analytics for Claude Code — dashboard access, CLI queries, cost analysis patterns, session tracking. Trigger on: "budi", "token cost", "how much did that cost", "session cost", "spending", "token analytics", "cost dashboard", "usage stats", or any question about Claude Code token consumption or spending patterns.From its SKILL.md

Install

npx -y skills add scoobydont-666/shared-claude-skills --skill budi-analytics

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
runs commandsInstructs the agent to run 8 commands, including `open http://127.0.0.1:<analytics-port>/dashboard` and 7 more.
fetches URLsInstructs the agent to fetch 1 URL, including http://127.0.0.1:<analytics-port>/dashboard.

SKILL.md

2.3 KB, 525 tokens by cl100k_base, as published. Nobody here has run it

budi Analytics

budi (WakaTime for Claude Code) tracks tokens, costs, and file activity per session.

Quick Reference

# Dashboard (web UI)
open http://127.0.0.1:<analytics-port>/dashboard

# CLI stats
budi stats                    # summary for current repo
budi stats --all              # all repos
budi stats --json             # machine-readable
budi sessions                 # list sessions
budi cost                     # cost breakdown
budi doctor                   # health check

Architecture

Daemon: budi-daemon on 127.0.0.1:<analytics-port> (started automatically)
DB: ~/.local/share/budi/repos/<repo-hash>/budi.db (SQLite, mode 600)
Hooks: 5 events in ~/.claude/settings.json (SessionStart, UserPromptSubmit, PostToolUse, SubagentStart, Stop)
Status line: live cost shown in Claude Code terminal

What It Tracks

Data	Source Hook	Stored?
Session start/end	SessionStart, Stop	Yes
Prompt text	UserPromptSubmit	Yes (privacy note: prompts in SQLite)
Token counts	UserPromptSubmit	Yes
Model used	UserPromptSubmit	Yes
Cost estimate	Computed from model + tokens	Yes
Files touched	PostToolUse (Read/Write/Edit/Glob)	Yes (paths only, not content)
Subagent spawns	SubagentStart	Yes

Coexistence with Our Hooks

budi hooks don't conflict with our custom hooks:

Our PreToolUse[Agent] (token-miser) → different event + matcher
Our SubagentStop (cost-tracker) → different event
budi uses HTTP hooks to daemon, ours use command hooks

Cost Model

budi uses Anthropic's published pricing. Same rates as token-miser:

Haiku 4.5: $1/$5 per MTok (in/out)
Sonnet 4.6: $3/$15 per MTok
Opus 4.6: $5/$25 per MTok

Troubleshooting

budi doctor                   # check hooks, daemon, config
systemctl --user status budi-daemon  # if daemon isn't running
budi-daemon &                 # start manually

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.