Documentation Index
Fetch the complete documentation index at: https://docs.tryhoard.com/llms.txt
Use this file to discover all available pages before exploring further.
Hoard ships with five permission modes per action kind. The defaults are conservative, most sellers should sit on them for a couple of weeks before changing anything. But once you know what kind of seller you are, you can pick a profile that matches your style.
This page is a decision tree. Find the description that sounds most like you, copy those settings, and don’t think about it again.
Refresher: the five modes
| Mode | When it asks you |
|---|
deny | Never. The action is fully blocked. |
always_ask | Every single call, regardless of risk. |
ask_for_mutations | Reads run free; any write asks first. (Default for most action kinds.) |
auto_low_risk | Low and medium risk run automatically; high and critical ask. |
auto_unless_critical | Only critical-risk calls ask. Power-user mode. |
”I just want it to read my data”
You want Claude or ChatGPT to answer questions about your store. Quick lookups, end-of-month summaries, “is this card priced right” gut checks. You don’t want it touching anything.
| Action kind | Mode |
|---|
All pricing.* write actions | deny |
All account.* write actions | deny |
sync.start | deny |
Everything in the read column | leave alone |
”I want suggestions but I’ll review each one”
This is the default profile. Sit here for two weeks. You’ll see what the assistant tends to suggest, you’ll click yes most of the time, you’ll catch the occasional thing that needs a second look.
| Action kind | Mode |
|---|
pricing.edit_rule | ask_for_mutations (default) |
pricing.mass_reprice | ask_for_mutations (default) |
account.settings | always_ask (default) |
sales.run | always_ask (default) |
”I trust the agent on small price tweaks”
You’ve watched the audit log for two weeks. You always approve low-risk pricing edits. The confirmations have started feeling like noise.
| Action kind | Mode |
|---|
pricing.edit_rule | auto_low_risk |
pricing.preview | auto_low_risk (probably already there) |
pricing.mass_reprice | ask_for_mutations |
account.settings | always_ask |
sales.run | always_ask |
”Power user, I’ll catch problems in the audit log”
You’re running multiple stores. You’re talking to the assistant several times a day. You’ve internalized that the floors and risk scorer are the real safety net, the per-call confirmation is the optional second check.
| Action kind | Mode |
|---|
pricing.edit_rule | auto_unless_critical |
pricing.mass_reprice | auto_unless_critical |
account.settings | always_ask (keep this) |
sales.run | ask_for_mutations |
”Never let it touch my account settings”
Whatever profile you pick above, you can pin account.settings to deny regardless. This is worth doing.
Account settings is where your sync mode, your timezone, your shipping adder, and your pricing template live. Nothing the assistant should be reaching for casually. Lock it down:
| Action kind | Mode |
|---|
account.settings | deny |
deny costs you nothing.
The 3-week trust ladder
If you’re not sure where to start, walk the ladder. Each step is one week. Each step changes exactly one grant. If anything weird happens, hold or drop back.
| Week | Change | Why |
|---|
| Week 1 | Sit on the defaults. Approve everything that looks right; reject anything that doesn’t. | You’re collecting data on the assistant’s actual behavior. |
| Week 2 | inventory.update_quantity → auto_low_risk. | Quantity edits are low-stakes (a single card’s listed quantity, fixable in one click). Earn trust on this namespace before loosening pricing. |
| Week 3 | pricing.edit_rule → auto_low_risk. | You’ve seen 10+ low-risk rule edits all be fine. Save yourself the clicks on those, keep the gate on bigger moves. |
auto_low_risk for pricing forever, you push further to auto_unless_critical, or you decide you actually like the confirmations and roll back. All three are legitimate.
For the full month-by-month version, see Going from cautious to confident.
When to roll a setting back
You loosened something. Now you’re seeing weirdness. How do you know it’s time to retighten?
The rough rule: two false positives in a week means roll back one step. A false positive is a commit you would not have approved if you’d been asked. One could be a fluke; two is a pattern.
Rolling back is the same UI as loosening, Settings → Assistants → permission grants. Move pricing.edit_rule from auto_low_risk back to ask_for_mutations. You’re back to a confirmation prompt on every edit.
The activity log is your forensic tool here. See Reading your agent activity log for how to spot patterns.
