Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Fixed credential declaration: using env/primaryEnv format for POSTWALL_API_KEY
What This Skill Does
PostWall is a security layer between AI agents and email. Incoming emails appear to agents only after a human approves them in the dashboard. Outgoing emails go through a draft queue and require manual approval before being sent.
Gives agents email access without granting unilateral send or read permissions, so hallucinated recipients or bad drafts never result in actual sent mail.
When to use it
- Having an agent draft a client update email for your review before it sends
- Letting an agent read support emails you have pre-approved in the dashboard
- Refining a draft with an agent before approving it from the dashboard
- Checking whether a submitted draft was approved, rejected, or sent
- Batch-marking processed inbox emails without fetching full content
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'Draft an email to sarah@example.com summarizing today's meeting and send me the approval link'
- 1Composes email body from meeting context
- 2Runs postwall draft --to sarah@example.com --subject 'Meeting Summary' --body '...' --json
- 3Parses approveUrl from the JSON response
- 4Returns the approval link to the user
Draft submitted with ID abc123. User receives a direct link to review and approve the email from the PostWall dashboard.
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.