More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Repair Agent Permissions, Approvals & Action Boundaries

    1

    Repair Agent Permissions, Approvals & Action Boundaries is a bounded ToolForge Labs workflow.

    $1

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Repair Agent Permissions, Approvals & Action Boundaries

    Repair Agent Permissions, Approvals & Action Boundaries

    Example session with this skill installed

    Our internal agent can read tickets, update records, and trigger a deployment tool. Read actions should be automatic, updates need team scope, and deployments require explicit human approval. We found traces where an old approval was reused after the deployment parameters changed. Repair the permission/approval boundary without granting broader access.

    • Read your context and instructions
    • Compiled the repair agent permissions

    The likely issue is a contract mismatch around agent permissions, least privilege, approval gates, delegated actions, scoped credentials, tool authority, escalation, session state, audit trails, and revocation. I would first define the invariant: Every executed action must be authorized at execution time for the exact principal, tenant, tool, operation and normalized parameters; approval cannot be treated as a reusable blanket permission. Then I would trace the smallest owning boundary, apply or propose a bounded repair/review step, and add deterministic normal, boundary, failure and replay tests. The workflow preserves existing authorization and business policy, user consent, separation of duties, credential boundaries, and destructive-action safeguards. If repository/workspace files are available, the findings can be mapped to exact paths and focused local validation; with partial evidence, the result stays an explicit diagnosis/evidence map with assumptions and no invented execution claims.

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

    About this skill

    Repair Agent Permissions, Approvals & Action Boundaries is a bounded ToolForge Labs workflow. Diagnose, review, and repair agent permissions, least privilege, approval gates, delegated actions, scoped credentials, tool authority, escalation, session state, audit trails, and revocation. Works from full repositories/workspaces or partial evidence. Produces bounded fixes or evidence gates with deterministic validation while preserving existing product/security policy and avoiding invented execution claims. It works from a full repository/workspace or partial evidence such as snippets, logs, configuration, traces, documents, tickets, and architecture notes. The skill separates verified facts from assumptions, avoids invented execution claims, and produces concrete repair/review steps with focused deterministic validation or traceable evidence.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.

    Requires clear definition of existing security policies. Bounded mocks are preferred over live production testing. Verification is limited to supplied context or repository access.

    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 buying a skill.

    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

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Use a clean branch or controlled review workspace. Supply repository/workspace files, logs, traces, configuration, documents, or evidence needed for the task. Review patches/findings before merge or operational action; keep production credentials and destructive production actions out of the workflow.

    Listed1 month ago
    Updated9 days ago

    What's inside

    Frequently Asked Questions

    This skill diagnoses and repairs security vulnerabilities in AI agents, such as over-privileged credentials, lack of human-in-the-loop approval gates, and insufficient audit logging. It ensures your agents operate within strict action boundaries without violating your existing security policies.

    The skill is framework-agnostic and works with any AI agent architecture that utilizes tool-calling, API credentials, or delegated actions. It can analyze full code repositories or partial evidence like log traces and configuration snippets.

    Unlike generic security tools, this skill produces deterministic validation steps and evidence gates rather than 'best guesses.' It strictly separates verified facts from assumptions, ensuring no 'hallucinated' execution claims are introduced into your codebase.

    You receive the full ToolForge Labs workflow designed to audit and patch access controls, including templates for scoped credentials, escalation protocols, and session state management. It also includes documentation on how to apply these fixes to various repository structures.

    Yes, the skill is designed to work effectively even with limited information such as architecture notes, system logs, or snippets of configuration files. It will identify what can be repaired immediately and where more evidence is required for a secure fix.

    New in Security