claude-permissions-optimizer

Auto-allow safe Bash commands in Claude Code by analyzing session history to identify frequently used, non-risky patterns and inserting them into your settings.json allowlist. This stops permission fatigue by automatically validating and adding only green rules.

Reduce permission prompts in Claude Code by scanning prior session transcripts to extract recurring safe Bash patterns. The tool normalizes commands, enforces count thresholds, and classifies them before updating your allowlist.

Yes, the allowlist optimizer works with other coding agents by analyzing data stored in ~/.config/opencode/ and extracting commands from prior Claude Code session transcripts, regardless of which agent initiated them.

Scope Bash command recommendations to a specific project by prompting the tool for project scope choices. It defaults to scanning all projects but honors custom filters so allowlist updates match your current work context.

Limitations include enforcing min-count thresholds to avoid over-reaching, and warning users when risky patterns surface. The tool deduplicates entries and validates JSON but requires running the extract-commands.mjs Node script first.

Yes, you need Node.js to run the bundled extract-commands.mjs script. This script normalizes Bash usage, classifies each pattern for safety, and prepares statistics before updating your settings.json allowlist.