Measuring AI Proficiency: Context Engineering as a Leading Indicator
Latest release v0.7.0 — v0.7.0 — Context Efficacy Proving Harness (--prove) · 1 Jun 2026
19 files for Claude Code, GitHub Copilot, Codex and OpenCode: improve-ai-context, agentic-workflow, context-surfing, customize-measurement and 15 more — 9,401 tokens loaded in every session.
.github/copilot-instructions.md A 1,459 tok AGENTS.md A 3,937 tok CLAUDE.md A 4,005 tok .claude/skills/agentic-workflow/SKILL.md A — .claude/skills/context-surfing/SKILL.md A 144 tok .claude/skills/customize-measurement/SKILL.md A 53 tok .claude/skills/eval-creator/SKILL.md A 67 tok .claude/skills/intent-framed-agent/SKILL.md A 54 tok .claude/skills/learning-aggregator/SKILL.md A 85 tok .claude/skills/measure-ai-proficiency/SKILL.md A 109 tok .claude/skills/plan-interview/SKILL.md A 108 tok .claude/skills/pre-flight-check/SKILL.md A 55 tok .claude/skills/simplify-and-harden/SKILL.md A 71 tok .claude/skills/use-agent-factory/SKILL.md A 170 tok .claude/skills/verify-gate/SKILL.md A 76 tok .claude/agents/improve-ai-context.agent.md A 39 tok .github/agents/create-agentic-workflow.agent.md A 29 tok .github/agents/debug-agentic-workflow.agent.md A 22 tok .github/agents/upgrade-agentic-workflows.md A 17 tok These files are pskoett/measuring-ai-proficiency's own configuration — they tell Claude Code, GitHub Copilot, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.