{"skill":{"slug":"tokenkiller","displayName":"TokenKiller","summary":"Reduces token usage across multi-skill agent workflows (search, coding, debugging, testing, docs) using budgets, gating, progressive disclosure, and deduped...","tags":{"agent":"1.0.1","budget":"1.0.1","cost":"1.0.1","debugging":"1.0.1","dedupe":"1.0.1","latest":"1.0.1","progressive-disclosure":"1.0.1","search":"1.0.1","tokens":"1.0.1"},"stats":{"comments":0,"downloads":318,"installsAllTime":2,"installsCurrent":2,"stars":1,"versions":2},"createdAt":1773216195729,"updatedAt":1777525848211},"latestVersion":{"version":"1.0.1","createdAt":1773225256595,"changelog":"- Introduced a detailed task complexity assessment to set budgets dynamically (simple/medium/complex).\n- Added soft warning and extension mechanism when approaching preset budgets, allowing more flexible handling.\n- Clarified scenarios for full content (L3) read access and added explicit “L3 pull” decision flow.\n- Defined multi-skill agent collaboration and clarified TokenKiller's priority as a constraint layer.\n- Added explicit self-check guidelines to minimize high token-consumption behaviors during execution.\n- Documentation was translated and expanded to English with new sections and improved structure.","license":"MIT-0"},"metadata":null,"owner":{"handle":"coderangerx","userId":"s17bc8a36bshpzjsf8ndjkv0x185r1s6","displayName":"CodeRanger","image":"https://avatars.githubusercontent.com/u/18438342?v=4"},"moderation":null}