Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    allowlist builder

    1

    Automate AI agent shell permissions by generating safe allowlists from your command history.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    allowlist builder

    allowlist builder

    Example session with this skill installed

    I'm tired of approving git status and npm test in Claude and Gemini. Check my history from the last 30 days and set up an allowlist for them.

    • Read your context and instructions
    • Compiled the allowlist builder
    • Generated the document

    I have analyzed your command history from the last 30 days across Claude Code and Gemini CLI. I found 42 instances of read-only git commands and 15 instances of project test commands that you consistently approved.

    Allowlist Proposal

    PrefixTierUsesAgents
    git statusread-only28Claude, Gemini
    git logread-only14Claude, Gemini
    npm testproject12Claude
    lsread-only9Gemini

    I have prepared the configuration files. For Claude Code, I will target the project scope (.claude/settings.json). For Gemini, I will update the global user policy.

    
    # Applying the allowlist rules...
    # Backup created: ~/.allowlist-builder-backups/2023-10-27-1430/
    # Claude Code: Added 3 rules to .claude/settings.json
    # Gemini CLI: Wrote ~/.gemini/policies/agent-allowlist.toml
    # Verification: PASS
    

    Next steps

    1. Restart your Claude Code session and Gemini CLI for the new policies to take effect.
    2. Review proposal.html in the skill folder if you want to see the specific high-risk commands that were excluded.
    3. To undo these changes, run the restore script pointing to the backup folder listed above.

    proposal.html

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop manually approving the same safe shell commands every time you use an AI coding agent. This skill analyzes your command history across Claude Code, Codex CLI, and Gemini CLI to identify safe, high-frequency patterns. It distinguishes between strictly read-only commands (like git status or ls) and project-specific execution commands (like npm test), allowing you to automate permissions for the former while maintaining strict oversight of the latter.

    What it does

    • History analysis scans recent agent logs to find frequently approved shell commands.
    • Tiered classification sorts commands into read-only, project-specific, or high-risk categories.
    • Multi-agent sync generates native configuration rules for Claude Code, Codex, and Gemini CLI simultaneously.
    • Safety filtering automatically excludes commands with dangerous flags like -delete, -exec, or --output.
    • Automated backups creates timestamped restores of all configuration files before making changes.

    How it works

    1. Plan by running the analysis script to see which commands are candidates for your allowlist.
    2. Review the proposed tiers and select which agents and project scopes you want to update.
    3. Apply the configuration to write native permission files and backup existing settings.
    4. Verify the new rules against the agent's internal policy checkers to ensure they parse correctly.

    Frameworks & tools

    Built for Node.js 18+. Integrates with Claude Code permissions, Codex CLI prefix rules, and Gemini CLI TOML policies.

    Why this beats prompting it yourself

    Manually editing agent policy files is error-prone and requires knowing three different regex and configuration formats. This skill ensures your allowlist is derived from your actual usage history while enforcing safety guardrails that prevent accidental automation of destructive commands.

    Use cases

    • Automate git and ls approvals across all local AI coding agents.
    • Share project-specific test and lint allowlists with a team via .claude/settings.json.
    • Reduce interruptive permission prompts during long refactoring sessions.
    • Audit which commands your AI agents are actually running most frequently.

    Known limitations

    Requires Node.js 18+ and local filesystem access to agent config directories. Cannot automate Cursor's allowlist due to UI-only settings, but provides the list for manual entry.

    Requires Node.js 18+. Cannot write to Cursor settings directly (UI-only). Ignores commands with redirections or substitutions for safety.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    Requires

    macOS
    Node.js

    Permissions

    Terminal / Shell
    Listed1 month ago

    What's inside

    Frequently Asked Questions

    The Allowlist Builder automates the permission-granting process by analyzing your command history to identify safe, repetitive patterns, eliminating the need to manually approve every 'ls' or 'git status' request while you work.

    It currently supports native configuration for Claude Code, Codex CLI, and Gemini CLI; for tools with UI-only settings like Cursor, it generates a formatted list that you can copy and paste manually.

    The skill uses a tiered classification system that automatically flags and excludes destructive flags (like --delete) and keeps project-specific execution commands separate from global read-only commands for maximum oversight.

    No, the skill operates locally on your machine by scanning your agent logs and writing to your local configuration files; it does not send your command history or code to any external servers.

    You will receive the full toolset compatible with Node.js 18+, including the analysis script, the multi-agent configuration generator, and the automated backup utility to protect your existing settings.

    New in Workflow Automation