overview

overview is a command for Claude Code from Kastalien-Research/thoughtbox. It costs 0 tokens per session (2,708 once invoked), scanned A, original, MIT.

An explanation of Hypothesis-Driven Development (HDD), a software workflow that treats Architecture Decision Records (ADRs) as the main source of truth and tests predictions before accepting or rejecting them.

In plain words
What is it for?
Use it to learn the HDD approach, its five phases, and the supporting loops used to explore, author, refine, verify, and coordinate work.
Why use it?
It gives developers the full process and explains how research, planning, implementation, refinement, and verification fit together.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/kastalien-research/thoughtbox/overview
Clone the repo
git clone --depth 1 https://github.com/Kastalien-Research/thoughtbox

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for overview

README.md
[![agentmods](https://agentmods.dev/badge/commands/kastalien-research/thoughtbox/overview.svg)](https://agentmods.dev/commands/kastalien-research/thoughtbox/overview)
Your own site
<a href="https://agentmods.dev/commands/kastalien-research/thoughtbox/overview"><img src="https://agentmods.dev/badge/commands/kastalien-research/thoughtbox/overview.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,708 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.02708
Opus 5 $0.00000 $0.01354
Sonnet 5 $0.00000 $0.00542
Haiku 4.5 $0.00000 $0.00271

Measured yesterday against content hash d184ff0c599e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

overview scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

apps/web/.claude/commands/hdd/overview.md · 373 lines

How it starts

The opening of the file, as written. The whole thing — 373 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Hypothesis-Driven Development: Complete Overview

What is Hypothesis-Driven Development?

Hypothesis-driven development (HDD) is a workflow that treats ADRs (Architecture Decision Records) as the source of truth rather than code. Before writing any code, we form testable hypotheses about what we expect to happen, document them in staging ADRs, implement, validate, and then either accept or reject based on whether reality matched our predictions.

Loop Building Blocks

The HDD workflow composes OODA loops across all 5 phases. See the unified /hdd command for the full loop architecture:

Category Loops Reference
Exploration problem-space, codebase-discovery, domain-research @loops/exploration/
Authoring spec-drafting, code-generation @loops/authoring/
Refinement requirement-quality, code-quality, consistency-check @loops/refinement/
Verification acceptance-gate, integration-test @loops/verification/
Orchestration queue-processor, spiral-detector, dependency-resolver @loops/orchestration/

See @loops/README.md for the full loop library. See .claude/commands/hdd/hdd.md for the executable unified command.


The Problem HDD Solves

Traditional development treats code as the source of truth, which creates problems:

  1. Agent collaboration is hard: Agents must infer intent from implementation details
  2. Change reasoning is opaque: Why decisions were made is lost after implementation
  3. Failed attempts are invisible: Teams repeat mistakes because failures aren't documented
  4. Architectural drift: The "why" behind decisions erodes over time
  5. Refactoring is risky: Without documented invariants, unclear what can change

Core Philosophy

Code is an Implementation Artifact

Code shows how something is done. ADRs explain:

  • Why we chose this approach
  • What alternatives we considered
  • What outcomes we predicted
  • Whether reality matched predictions

Read the full file on GitHub · 373 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 373 lines · 0 tokens per session scan A d184ff0c599e

Subscribe to this mod's changes

overview is a command published in the GitHub repository Kastalien-Research/thoughtbox (64 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,708 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.