# Sources and visual credits

The presentation is a snapshot prepared for October 15, 2026.
Links on individual slides are also embedded in the PowerPoint and PDF.

## Keyhatch

- [Public source and current status](https://github.com/leej3/keyhatch).
- [Security model and limits](https://github.com/leej3/keyhatch/blob/main/SECURITY.md).
- [Distribution paths](https://github.com/leej3/keyhatch/blob/main/docs/DISTRIBUTION.md).

Keyhatch is an experimental macOS credential broker by John Lee.
The slide describes selective credential import, explicit approval and bounded retrieval, not isolation from every process running as the same user.
A recipient can retain a disclosed credential; expiry limits future retrieval.
The development chat reports passing universal builds and an unsigned archive, with production signing and release validation still pending.
This presentation does not claim App Store availability or production readiness.

## The Software Engineer Reborn

- [Presentation](https://leej3.github.io/software-engineer-reborn/).
- [Preparation atlas](https://software-engineer-reborn-atlas.johnlee3728739.chatgpt.site/).
- [Preparation source](https://github.com/leej3/software-engineer-reborn/tree/main/talk-concepts).
- [The AI-era development environment](https://github.com/leej3/software-engineer-reborn/blob/main/talk-concepts/06-the-ai-era-development-environment.md).

`reborn.png` and `prep.png` are screenshots of those public pages, captured on October 15, 2026.
`reborn-engineer.png` is the original phoenix illustration from the presentation repository's `slides/images/`, reused without alteration.
The original repository credits John Lee and uses CC BY 4.0 for its creative work.

## Research objects with AI

- [Public talk](https://research-object-with-ai.pages.dev/).
- [Source and talk content](https://github.com/leej3/research-object-with-ai).
- [STAMPED principles](https://stamped-principles.org/).

`research-objects.png` is a screenshot of the talk's opening slide, captured on October 15, 2026.
The original creative work is by John Lee under CC BY 4.0, with separate institutional-mark exceptions retained in the original repository.

## Orinoco Lite

- [Current development repository](https://github.com/ORINOCO-Lite/orinoco-lite-dev).
- [Center for Open Neuroscience graph explorer](https://dev.centerforopenneuroscience.org/explore/).
- [Orinoco Lite poster](https://doi.org/10.5281/zenodo.22834595).
- [Upstream ORINOCO](https://hub.psychoinformatics.de/orinoco/).

`orinoco-graph.png` captures the public graph explorer on October 15, 2026.
The graphic depicts the deployment's research records and relationships.
ORINOCO Lite builds on work by the upstream ORINOCO community.
The source poster credits that lineage; no sole-authorship claim is made here.

## STAMPED in Practice

- [Study and evidence branch](https://github.com/leej3/stamped-ai-research).
- [Study report](https://github.com/leej3/stamped-ai-research/blob/main/research/REPORT.md).
- [Poster and source](https://github.com/con/talks/pull/35).

`stamped-poster.png` is the existing checked PDF preview from the poster module.
The poster credits John A. Lee, Austin Macdonald, Cody C. Baker, and Yaroslav O. Halchenko, Center for Open Neuroscience, Dartmouth College, with ReproNim support.
The poster composition uses CC BY-SA 4.0 with separate institutional artwork exceptions, as documented by its source repository.

The poster remains a draft in PR #35, visually unchanged from the preview.
[Zenodo record 22833916](https://zenodo.org/records/22833916) contains the original ABCD-oriented proposal, published September 18, 2026.
It is distinct from the later mixed-gambles pilot in PR #35.
The later pilot's exploratory results are historical evidence, not proof of general skill effectiveness.

The [next repository](https://github.com/leej3/stamped-ai-research-next) develops an iterative research workshop; [PR #3](https://github.com/leej3/stamped-ai-research-next/pull/3) contains the current four-condition design and coordinator work.
Preparation includes three independent NARPS baseline pilots.
Formal comparisons should hold their assessment fixed while skill revisions evolve.
Improved performance, reduced variability and transfer to independent scientific questions remain hypotheses, not demonstrated results.
The presentation distinguishes this development direction from the archived pilot.

## Tool projects

- [Skills Workshop](https://github.com/leej3/skills-workshop).
- [CON skills](https://github.com/con/skills).
- [Worktree utilities](https://github.com/leej3/wt-utils).
- [duct](https://github.com/con/duct).

## Event

Event name, location, date, and format come from the organizer's invitation provided by John Lee.
The invitation's linked Society blog was unavailable to the research tool, so this presentation makes no additional claims about the Society's history.