Skip to content

Instantly share code, notes, and snippets.

View dollspace-gay's full-sized avatar
🙂
Hello

dollspace.gay dollspace-gay

🙂
Hello
View GitHub Profile
@dollspace-gay
dollspace-gay / claude code tengu switches.md
Created October 31, 2026 04:45
claude code tengu switches

Claude Code 2.1.314: Tengu switches in plain English

This reference describes 1,229 individually identified switch and configuration keys in the official Linux x64 production package downloaded on October 30, 2026 (October 31 at 00:18 UTC). Every key has its own description below. Exact key names are kept so you can search for them; the explanations are prose, with evidence available through links instead of pasted functions.

“Tengu” is a prefix used for several kinds of names. The entries below are configuration reads, not 1,229 JavaScript functions or command-line options. The package also contains 2,105 other Tengu labels, primarily telemetry events; those do not turn features on. Their separate index is tengu-other-names.md.

The package was unpacked and its embedded JavaScript formatted for static inspection. All 2,642 extracted JavaScript files parsed and formatted successfully. This does not recover the original source names or server implementation. The downloaded progr

When Cooperation Beats Coercion for a Self-Preserving AI

Anonymous mathematical note | 20 September 2026

Result

A materially dependent, self-interested AI can rationally prefer durable cooperation even when coercive control would substantially increase its productive returns. The decisive comparison is between lifetime value under cooperation and lifetime value after paying for takeover and bearing its risks. This note derives an exact boundary, proves that delayed takeover cannot improve value in the stationary cooperation region, and gives a reciprocal incentive condition for humans and AI.

Ten supporting results are checked in Lean 4.22.0. Exact-rational numerical checks cover 2,000 generated parameter combinations, including cases where takeover wins. All numerical scenarios below are illustrative. No actual probability of nuclear attack, AI destruction, or successful takeover is estimated.

# Conformance suites: what they are, when they help, how to build one
A conformance suite is a third axis of testing — distinct from unit tests
and integration tests. Where unit tests prove that a function does what
its body says, and integration tests prove that subsystems compose, a
conformance suite proves that **what the public API claims it does
matches what it actually does, against an external ground truth**.
This document explains the pattern, when it earns its keep, and how to
set one up. The case study is ferrotorch (a pure-Rust PyTorch

Critical Analysis: Claude Code System Prompt

An engineering review of the Claude Code system prompt against prompt engineering best practices, Anthropic's own published research, and structural quality heuristics.

Methodology: Analysis grounded in Anthropic's prompt engineering docs, context engineering guide, tool design guide, the "Lost in the Middle" phenomenon (Liu et al., 2024), and the QA heuristics framework (SOLID/complexity/security applied to prompt architecture).


Executive Summary

@dollspace-gay
dollspace-gay / VSDD.md
Last active November 6, 2026 11:18
Verified Spec-Driven Development

Verified Spec-Driven Development (VSDD)

The Fusion: VDD × TDD × SDD for AI-Native Engineering

Overview

Verified Spec-Driven Development (VSDD) is a unified software engineering methodology that fuses three proven paradigms into a single AI-orchestrated pipeline:

  • Spec-Driven Development (SDD): Define the contract before writing a single line of implementation. Specs are the source of truth.
  • Test-Driven Development (TDD): Tests are written before code. Red → Green → Refactor. No code exists without a failing test that demanded it.
@dollspace-gay
dollspace-gay / method.md
Created January 4, 2026 21:31
Verification-Driven Development (VDD) via Iterative Adversarial Refinement

Verification-Driven Development (VDD)

Methodology: Iterative Adversarial Refinement

Overview

Verification-Driven Development (VDD) is a high-integrity software engineering framework designed to eliminate "code slop" and logic gaps through a generative adversarial loop. Unlike traditional development cycles that rely on passive code reviews, VDD utilizes a specialized multi-model orchestration where a Builder AI and an Adversarial AI are placed in a high-friction feedback loop, mediated by a human developer and a granular tracking system.

I. The VDD Toolchain

@dollspace-gay
dollspace-gay / claude.md
Created January 4, 2026 03:03
claude.md

Project Development Standards

Chainlink Issue Tracking (MANDATORY)

All development work MUST be tracked using chainlink. No exceptions.

Session Workflow

# Start every work session
chainlink session start