agent-context-audit

To audit AGENTS.md instruction density, the skill evaluates root and nested files to measure depth, identify duplicated guidance, and detect coverage gaps across multi-level agent trees. It constrains results to a defined schema and provides actionable improvement guidance.

AGENTS.md nested depth measures how instruction files are structured across multi-level agent trees. High nestedness can create context bottlenecks, while missing coverage leaves agents without necessary guidance, making structural auditing essential for maintaining context quality.

Finding duplicated guidance and coverage gaps in AGENTS.md files requires auditing root and nested structures against a defined schema. The audit identifies redundant instructions across multi-level trees and reports missing coverage areas to prioritize agent-context improvements.

Yes, you can audit multi-level AGENTS.md agent trees to identify structure issues. The audit scopes instruction density, nested depth, duplication, and coverage gaps across root and nested files, constraining evaluations to a defined AGENTS.md schema for accurate reporting.

Common AGENTS.md structure bottlenecks include excessive nested depth, duplicated guidance, and coverage gaps across agent trees. The audit identifies these issues by evaluating instruction density against a defined schema and provides prioritized, actionable improvement guidance to resolve them.