What problem does it solve? LLM agents degrade unpredictably as context grows: critical information gets lost in the middle of long prompts, hallucinations poison downstream reasoning, irrelevant documents dilute attention, and contradictory sources cause silent failures. This Skill provides detection signals, thresholds, and mitigation strategies for each failure pattern. ## Core Features & Use Cases - Pattern Diagnosis: Identify five distinct degradation patterns (lost-in-middle, poisoning, distraction, confusion, clash) with specific detection signals for each. - Health Analysis Scripts: Run Python tools that measure attention distribution, detect poisoning indicators, and compute a composite context health score with recommendations. - Mitigation Framework: Apply the Write/Select/Compress/Isolate strategies matched to the active degradation pattern, with compaction triggers set before degradation cliffs. - Use Case: An agent's output quality drops after 40K tokens of conversation. Use this Skill to confirm lost-in-middle placement of critical instructions, then restructure context to move key constraints to attention-favored positions. ## Quick Start Analyze my agent's conversation context for degradation patterns and recommend whether to compact, truncate, or restructure it.
