Early platform · not a store SKU

Secondhand Decisions

Don't research it twice.

You already did the hard part of a decision. Sell that trail. Someone with the same constraints buys it — instead of starting from zero.

What's live vs next

Live now

  • This platform page
  • Buyer checkout — $7, working
  • Paste a messy dump → local packet preview
  • Agent-shaped doors (seller glance + buyer matches)
  • Build-your-own templates (T1–T3)
  • Example inventory — 8 labeled EXAMPLE packets

Coming next

  • Seller publish API (a listing that actually goes live)
  • Constraint matching against real inventory
  • Seller payouts — $3 per sale

The easy door

Agents first

This is the path that has to feel almost boringly simple. Humans without an agent still have a door below — same packet, same $7. No blank-form main path.

Seller · with agent

  1. 1You already decided something (laptop, insurance, city, tool…).
  2. 2Tell your agent: “Packet this decision.”
  3. 3Agent pulls notes / history / rejects / winner → one packet.
  4. 4You glance. One tap to publish when listings go live.

Drop the trail you already have. Not a form.

Early inventory is copy the packet. Email send to the Fort is next via Fort Mail — never Gmail. Listing fee is $0. Publish + $3 payouts are next.

Buyer · with agent

  1. 1Tell your agent your constraints (budget, must-haves, dealbreakers).
  2. 2Agent finds overlapping packets.
  3. 3You see 1–3 matches, not a junkyard.
  4. 4One tap: buy the trail. Skip the rabbit hole.

Early matching is local example inventory — the shape, not a live catalog.

Checkout is live. A full catalog of seller packets is next — this $7 rail is the buyer door.

Still open

No agent? Still short.

Human path · slower, still short

Paste a messy dump (notes, screenshot text, “I looked at A/B/C and picked B because…”). We standardize it here in the browser. Same packet shape. Same $7. Not as frictionless as the agent door — but not a form hell.

Copy, don't fill twelve fields

Build your own

Three templates. Paste into your notes or hand them to an agent. The graveyard (what you rejected) is half the value — don't skip it.

T1 · blank packet schema

Blank packet

Fill the six fields + 2–4 tags. Missing outcome is allowed — mark it, don't invent it.

SECONDHAND DECISIONS PACKET
Tags: [2–4 tags]
Constraints:
Rejected:
Winner:
Outcome:
Would do differently:

T2 · agent prompt

Packet this decision

The same sentence as the seller door. Copy it into any agent with your messy notes underneath.

Packet this decision.

You already decided something. Turn my messy notes into one Secondhand Decisions packet using this schema:

- tags (2–4)
- constraints
- rejected (the graveyard — keep each reject concrete: what + why)
- winner
- outcome (after living with it)
- would do differently

Rules: keep rejects concrete. Do not invent outcomes I did not state — mark missing fields instead of guessing. Do not add contact addresses. Return the packet as plain text I can copy.

T3 · category shells

Start from the kind of decision

Four shells. Copy the one that matches, then fill. Still not a form.

Buy under $budget

BUY UNDER $BUDGET (hardware / gear)
Tags: [thing], under $[budget], [must-have]
Constraints: Budget ≤ $___ · must-haves · dealbreakers · used vs new
Rejected: [option] — [why] · [option] — [why]
Winner: [what you got] — [why it cleared the constraints]
Outcome: [weeks/months later: still happy? what broke?]
Would do differently:

Pick a tool

PICK A TOOL (software / SaaS / agent stack)
Tags: tool, [job-to-be-done], [stack]
Constraints: Job to be done · budget · who has to operate it · lock-in you will not accept
Rejected: [tool] — [why] · [tool] — [why]
Winner: [tool] — [why it won]
Outcome: [after using it for real work]
Would do differently:

Pick a place

PICK A PLACE (city / neighborhood / move)
Tags: city, move, [jobs|quiet|rent]
Constraints: Work shape · rent ceiling · noise · heat/cold · groceries/transit
Rejected: [place] — [why] · [place] — [why]
Winner: [place] — [why it cleared the constraints]
Outcome: [after a season, not a weekend]
Would do differently:

Pick insurance

PICK INSURANCE (boring coverage)
Tags: insurance, [renters|auto|health], under $[n]/mo
Constraints: Coverage you actually need · monthly ceiling · no app-only login hell
Rejected: [policy] — [why] · [policy] — [why]
Winner: [plain policy] — [why boring won]
Outcome: [a claim, or a quiet year]
Would do differently:

Not live seller listings

Example inventory

Eight packets so you can see the shape with real-looking detail. Each one is marked Example only. Buyer matching above uses this same set locally. Lived packets from actual sellers replace these when publish ships.

Money · starter

$7 splits clean

$7Buyer pays
$3Seller gets
$4Platform keeps
$0Listing fee early

Scarcity early is inventory, not fees. Operator: The Fort That Holds LLC.

A room of the Fort — not a Grand Bazaar product page. Store may point here later. Terms & ops.