See your project from your agent’s point of view.
Pick any file or folder and agentpov shows what your coding agent works with there: the instructions and memory it reads, the permission rules and hooks that apply, the skills, subagents and MCP servers it can reach, and where each one came from.
macOS 13+ · Apple Silicon & Intel
How it works
Instructions
What the agent is told
Every instruction and memory file that reaches this path, from org-wide policy down to the folder itself, in the order the agent reads them.
Permissions
What it may do
Which allow, ask or deny rule decides each tool call on this path, and which rules a higher layer overrides.
Hooks & tools
What runs alongside it
The hooks that fire when the agent edits here, plus the skills, subagents and MCP servers it can reach, grouped by where each is defined.