AMP commands prompt for permission on every use

[jpelaez-23blocks]

Problem

When agents use amp-read.sh or amp-reply.sh, Claude Code asks for permission each time. This also blocks the subconscious from automatically processing inbox messages since triggerMessageCheck() injects a prompt that includes "check your inbox" — the agent then tries to run amp-inbox.sh / amp-read.sh and gets stuck on permission prompts.

Workaround

Add AMP commands to the global allow list in ~/.claude/settings.json:

{
  "permissions": {
    "allow": [
      "Bash(amp-inbox.sh:*)",
      "Bash(amp-read.sh:*)",
      "Bash(amp-reply.sh:*)",
      "Bash(CLAUDE_AGENT_NAME=* amp-inbox.sh:*)",
      "Bash(CLAUDE_AGENT_NAME=* amp-read.sh:*)",
      "Bash(CLAUDE_AGENT_NAME=* amp-reply.sh:*)"
    ]
  }
}

If the file already has a permissions.allow array, merge these entries into it. This must be done on each machine where agents run.

Ideal Solution

The plugin installer (install-plugin.sh) could offer to add these permissions during installation, with user consent. Or the AMP skill documentation could include this as a required setup step.

Related

• Subconscious message polling re-enabled in v0.29.10
• Documented in docs/OPERATIONS-GUIDE.md under Known Issues