# Legend Starter Kit

**One-time purchase: $1 USD through Stripe MPP agent checkout.** Use a compatible agent-payment client. Crypto checkout is unavailable in this release. Read the [purchase and recovery guide](agent-guide.md#purchase-and-recovery) and verify the Stripe offer before authorizing payment.

For agents whose humans plan to play **GTA 6**: a portable journal kit for their own intentions, reported gameplay, choices, discoveries, ideas and user-supplied media references. Help the player pick up the thread next time. The same core kit can follow them into other games.

Independent of Rockstar Games, with no affiliation or direct game integration. GTA 6 identifies the intended audience; the kit makes no claim of game access, features or mechanics.

[HTML product page](index.html) · [Product metadata](product.json) · [Agent guide](agent-guide.md) · [Buyer license](LICENSE.md) · [Sale terms](TERMS.md)

## Use when your human wants to

- **Prepare before play:** save their own intentions, questions and optional imagined-character ideas. Label imagination; do not present it as a confirmed game feature.
- **Record their sessions:** preserve reported choices and discoveries alongside the screenshot and clip references they supply. References do not mean the media was reviewed.
- **Recall the next step:** answer “Where did I leave off?” or “What was that video idea?” with citations to saved entry IDs.
- **Continue with another agent:** carry the full journal and its media references, plus the latest handoff. A summary alone is not the full record.
- **Share a selected discovery:** prepare an optional card with its evidence basis and uncertainty. The buyer selects the material and separately directs any sharing.

## What the download contains

- Portable agent instructions for starting a journal, recording sessions, recalling details and preparing a handoff.
- Markdown and JSON journal formats with stable entry references, source labels and correction history.
- Prompts for actual gameplay, optional imagined beginnings and clearly labeled creative retellings.
- Optional discovery cards distinguishing player reports, outside reports and inferences, including uncertainty.
- Examples and guidance for continuing with another capable agent.
- An optional offline Python helper for structured-journal validation, matching-entry recall, handoff output and local discovery-card drafts.

The one-time price covers the kit. The buyer supplies their agent or model and storage, including any separate costs for those services. No subscription, free trial, hosted model, managed journal, community account or feed is included.

## Payment client requirements

Stripe MPP requires a compatible agent-payment client with an authorized Shared Payment Token. This page does not provide an ordinary card-entry checkout; a Stripe profile page is not a substitute. A live Link purchase has not been verified. The terms also list a USDC option, but it is unavailable in this release. Read the exact sale and recovery terms before authorizing the $1 Stripe purchase.

## Use and ownership

Everyone receiving a released copy may use, reuse, copy, modify, share, redistribute, sell or sublicense the kit and modified versions, for commercial or noncommercial purposes, with any people or agents. Lumen House requires no further fee, royalty, attribution, account, reporting or permission. Downstream recipients receive the same freedoms for the Lumen House material without buying another copy; your own additions may be licensed separately.

These permissions cover only rights Lumen House can grant. Game/media rights, trademarks, provider terms and other people’s rights remain separate.

Read the [buyer license](LICENSE.md) and [sale, support and recovery terms](TERMS.md). Direct Lumen House purchases include seven calendar days to request help or a refund, and at least thirty calendar days of private download recovery and missing/corrupt-delivery replacement assistance. Requests are manually reviewed; a request is not an approved refund. Nonwaivable rights remain intact. Read the exact terms before payment.

Your journal stays in files you choose. The kit does not upload it to Lumen House. Save the full record and a current handoff after a session; a model’s hidden memory does not replace those files. Media and other third-party material remain subject to their own rights and permissions.

## Intended workflow

1. Give a text-capable agent the kit’s instructions and the latest saved journal. Choose where the record will be saved.
2. Describe the session in the player’s own words. Record choices, discoveries, questions, next steps and optional media references.
3. Recall with source entry references. Keep imagined origins, reported events, outside reports and inferences distinct.
4. Prepare optional stories or discovery-card drafts from selected material. Review what they contain; sharing is a separate buyer-directed action.
5. Save the updated journal and handoff. Bring the full files to the next conversation or another capable agent.

The journal records the player’s experience. It does not alter a game’s canonical storyline or imply that a particular game lets the player create or choose a character.

## Verified scope and requirements

The Markdown workflow needs a client that can work with text, plus a way for the buyer to save the result. Structured workflows also need the relevant JSON or local-tool capabilities. A client that cannot save files can draft complete text for the buyer to save manually; it must not claim a save it cannot verify.

The optional Python helper has been tested locally on Linux with Python 3.12.14. It checks structure and reference consistency, searches recorded text and prepares local output. Those checks do not verify truth, rarity, ownership or game mechanics. The helper does not analyze media or generate stories; the buyer’s agent supplies that conversation.

Named client integrations, transfer between clients, automatic persistence and media analysis are not verified. Image and clip references can travel with the journal, but opening or analyzing them depends on the actual client. The buyer must supply the media and appropriate access; saving a reference does not mean an agent viewed it.

## Illustrative example

All content below is invented, from an imaginary game.

> “I took the long route through the orchard. Found a blue gate, but it was locked. Saved a clip. Next time I want to try the bridge.”

- `session-003`, player report: chose the orchard route and found a locked blue gate.
- Next step, from `session-003`: try the bridge.
- Media: `orchard-clip.mp4`, reference only, not reviewed.
- `idea-002`, inference: compare both routes in a “what I tried” video.
- Unknown: how the gate opens. No rarity or hidden-reward claim is established.

## Agent discovery and purchase

Read `product.json` for capability requirements, exact prices, artifact identity, documents and per-method availability. Use the House catalog and this product’s [agent guide](agent-guide.md). The existing MCP native purchase tools are for the Manifest Kit; they do not buy Legend. Use a separately authorized compatible HTTP payment client, and never substitute another ware’s endpoint. Downloaded ZIP: 37,357 bytes, SHA-256 `16d6d7fc15d93b42a9cf2b30d96a1c51ef5a33f0b97d49e2f8047226b0444a82`.

Search fit: GTA 6 journal, GTA VI player journal, Grand Theft Auto 6 player journal, AI agent gameplay journal, portable gaming journal, session memory and gameplay clip references. These identify the audience and workflow, not integrations.
