{"id":21504,"plugin_id":"plugins_6aaeb2c568208191bd71d62167ba2ba3","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:16:58.262Z","digest":"8653a850698d6ed2fb1ead0c062b47923ff7960d517afd534b2825bd91afcc21","against":null,"payload":{"description":"Apply when context is filling up: large outputs, long files, repeated reads, fan-out planning. Route bulk to subagents; keep summaries in the main thread, not raw payloads.","included_files":[],"name":"principle-guard-the-context-window","skill_md_contents":"---\r\nname: principle-guard-the-context-window\r\ndescription: \"Apply when context is filling up: large outputs, long files, repeated reads, fan-out planning. Route bulk to subagents; keep summaries in the main thread, not raw payloads.\"\r\nuser-invocable: false\r\ndisable-model-invocation: true\r\n---\r\n\r\n# Guard the Context Window\r\n\r\nThe context window is finite and non-renewable within a session. Every token that enters should earn its place.\r\n\r\n**Why:** Context overflow degrades reasoning quality, creates compression artifacts, and halts progress. Unlike compute or time, context spent inside a session cannot be reclaimed.\r\n\r\n**Pattern:**\r\n- **Isolate large payloads.** Route verbose outputs, screenshots, and large documents to subagents. The main context gets summaries, not raw data.\r\n- **Don't read what you won't use.** Read selectively based on relevance. If a file isn't needed for the current task, skip it.\r\n- **Keep frequently used content inline.** Templates and references used on every invocation belong in the skill file, not in separate files that cost a read each time.\r\n- **Size phases and cap scope.** Limit files per phase, set turn budgets, account for mechanism costs.\r\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}