{"version":"2.1.121","anchor":"cost-attribution-by-agent-skill-and-plugin","canonical_anchor":"cost-attribution-by-agent-skill-and-plugin","heading":"Cost Attribution by Agent, Skill, and Plugin","tier":null,"area":null,"scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.121\/e\/cost-attribution-by-agent-skill-and-plugin","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.121","markdown":"### Cost Attribution by Agent, Skill, and Plugin\n\n**What**\n\nThe usage\/cost display now breaks down spending by agent, skill, and plugin, showing what percentage of usage each component consumed.\n\n**Details**\n\n- API responses now carry `attributionAgent`, `attributionSkill`, and `attributionPlugin` fields\n- Costs are accumulated per-request into `byAgent`, `bySkill`, and `byPlugin` maps\n- The cost view renders a \"Skills, subagents, and plugins\" section with percentage breakdown\n- Actionable tips appear alongside high-cost items:\n- Skills: \"Heavy skills can be scoped down or run with a cheaper model via skill frontmatter.\"\n- Plugins: \"Review what this plugin contributes \u2014 its agents, skills, and MCP tools all count toward your limit.\"\n- Shows \"No attribution data yet \u00b7 accumulates as you use Claude\" when no data exists\n\n**Evidence**\n\nAttribution tracking (search for `\"attributionAgent\"`, `\"attributionPlugin\"`, `\"Skills, subagents, and plugins\"`)"}