Credentials. Granting a contact's agent access to your calendar or other data is a real permission, so review each grant and revoke ones you no longer need.
Who it is forA person who runs a personal agent and wants a partner, friend or team agent to get specific access without open access.
The problemSharing a calendar or task list with another person's assistant usually means giving it everything. You want limits that you can read and change.
The outcomeYour agent applies seven access tiers in code before any tool runs, and you get a one-line text saying who asked for what.
What you need first
- Node 22.19+ and pnpm 10
- An Anthropic API key (or another Pi provider key)
- An Inkbox admin API key from inkbox.ai for phone, email and agent-to-agent endpoints
- A Maritime API key from maritime.sh if you deploy to a VM
- A Stripe Link OAuth client if you want the agent to pay (review takes about 3 to 5 days)
Steps
- Set up your agent with
pnpm instinct init --name "Maria" --phone +14155550100 --email [email protected] --handle maria-instinct - Read the seven tiers in docs/PERMISSIONS.md and the agent-to-agent protocol in docs/PROTOCOL.md
- Invite a contact through the network features described in packages/network
- Add a grant in plain English such as "Sam can book us dinner this week"
- Check the one-line text you receive when a contact asks for something
Prompt for your agent
In the open-instinct repo at <repo path>, read docs/PERMISSIONS.md and docs/PROTOCOL.md, then list the seven tiers and how a grant is written. Expected output: a short list of the tiers and one example grant. Verify that a stranger's request is answered with a no in the summary you produce.
Caveat. The README does not list the exact CLI commands for invitations and grants, so read packages/network and docs/PERMISSIONS.md before you invite anyone.
Get 5 new use cases every morning. Free, one short email, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Disclaimer
Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page