AI coding

Permissions and least privilege for agents

• A .agentignore for Your Agent: What to Hide and Leave Visible
• A File-Access Policy That Survives a Repo Move
• Auditing an Allowlist: Finding Rules No Agent Fires
• Guardrails for Legacy Interfaces: What an Agent May Not Touch
• One Allowlist for Tool Calls and File Reads
• Read-Only by Default: Agent Tasks Without Writes
• What an Agent May Do to a Database: Grants, Roles, Read-Only
• A Deny List That Outranks the Allowlist: Precedence and Tests
• A Deny Rule That Outranks the Manifest: Precedence and Tests
• A Registry Policy for Agent Package Installs
• Auditing an Agent's Global Install Script Allowlist
• Automating the Path Allowlist with Git Hooks and CI Checks
• Column-Level Grants for Agent Users: Only What a Task Needs
• File and Path Deny Rules for Agents: A Policy and Its Tests
• Grants and Read-Only Users: What an Agent May Do
• How Deny Rules Cost Tokens: Retries After a Refusal
• Migrating a Database with an Agent in the Loop: Who Runs What
• Path Deny Rules for Business Data an Agent Must Never Read
• Permission Tiers for Agent Writes: Designing a Roles Table
• Scoping Agent Permissions to Paths: Allowlists and Deny Rules
• Scoping an Agent to One Module: Path Rules and Read Boundaries
• Scoping Read Access to One Team's Business Data
• The Read-Only Contract an Agent Runs Before It Writes
• When a Dry Run Reads Sensitive Data
• When the Read Is the Risk: Capping Rows and Sensitive Fields
• Who May Change a System Prompt: A Roles and Approval Table
• Who May Edit Test Files and System Prompts