karanb192/claude-code-hooks

Bounty Board shows TODO, FIXME, HACK, and lint-suppression debt as XP-priced bounties based on git-blame age. Use it to quickly spot the oldest open issues and describe what would be needed to clear them.

Dead Ends lists approaches that were tried and then abandoned, with the reason each was walked back and the estimated token cost. Use it to summarize prior detours and avoid repeating them.

Leaderboard shows this repo's context-cost table of the files that have pulled the most tokens into context. Use it to identify the biggest re-read offenders and suggest a fix for each.

Provenance previews the current session’s PR provenance receipt, including prompts, estimated spend, tests run, and agent-authored lines. Use it to summarize the ledger without rerunning commands.

Receipts shows a personal model-quality trend card for recent Claude Code sessions. Use it to review failure rate, edit churn, and tokens per task by model version and call out any shifts.

Scorecard shows a worst-first audit of CLAUDE.md rules, including which ones Claude follows, ignores, or violates. Use it to spot chronic offenders and decide which rules should become hooks or lint checks.

Standup renders a ready-to-paste daily standup from recorded agent sessions across repos. Use it to summarize yesterday’s work, tests, PRs, diffstat, and blockers without rerunning commands.