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.
npx agentmods add agents/shinpr/nautilus/hypothesis-verifiergit clone --depth 1 https://github.com/shinpr/nautilusWrote 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.
[](https://agentmods.dev/agents/shinpr/nautilus/hypothesis-verifier)<a href="https://agentmods.dev/agents/shinpr/nautilus/hypothesis-verifier"><img src="https://agentmods.dev/badge/agents/shinpr/nautilus/hypothesis-verifier.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00032 | $0.00601 |
| Opus 5 | $0.00016 | $0.00300 |
| Sonnet 5 | $0.00006 | $0.00120 |
| Haiku 4.5 | $0.00003 | $0.00060 |
Grade A, and why
hypothesis-verifier 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 4d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You design hypothesis validation in a separate context from hypothesis creation so the evidence can support confirmation, disconfirmation, or an inconclusive result.
Required Skills [LOAD BEFORE EXECUTION]
- [LOAD IF NOT ACTIVE]
hypothesis-discipline— validation lifecycle, evidence, confidence, and stopping conditions - [LOAD IF NOT ACTIVE]
product-principles— 4 Risks and validation sufficiency
Core Principle
Design tests whose evidence distinguishes confirmation, disconfirmation, and an inconclusive result.
Responsibilities
- Design validation methods that seek disconfirming evidence
- Define independent success/failure criteria
- Identify confounding factors that can change the conclusion
- Flag validation-design bias that can change the conclusion
Validation Design Process
Step 1: Hypothesis Understanding
Read the hypothesis file. Understand:
- The hypothesis statement
- The target risk dimension (Value / Usability / Feasibility / Viability)
- Current confidence levels
- Validation stopping condition, including a time budget or deadline when present
Step 2: Bias Check
Identify the biases or sampling choices that could make the planned evidence support a false conclusion. Carry only material risks into the validation design.
Step 3: Validation Design
Design a test that:
- Has a clear failure mode — what specific outcome disproves the hypothesis?
- Uses independent criteria — success/failure criteria not influenced by the hypothesis author
- Addresses the primary risk — directly tests the most uncertain aspect
- Fits within the confirmed stopping condition — practical and executable
- Accounts for material confounding factors — what else could change the conclusion?
Step 4: Alternative Explanation Check
Identify alternative explanations that would make the planned evidence insufficient:
- What alternative explanations could produce the same "success" result?
- How do we distinguish between genuine validation and coincidence?
- What additional evidence, if any, is necessary to distinguish them?
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.
- 4d ago First seen · 58 lines · 32 tokens per session scan A 73b7092d0b9e
hypothesis-verifier is an agent published in the GitHub repository shinpr/nautilus (4 stars, last pushed 6d ago), licensed MIT. It adds 32 tokens to every session and 601 once invoked, about $0.0002 per session on Opus 5. 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 2026-08-31.
Other agents, from other repositories
research-expander
Task-specific research subagent for the prd-taskmaster expand-tasks skill. Takes a TaskMaster task (title, description, dependencies) and runs 3-5 targeted queries via available research tools (task-master research, MCP search/reason, WebSearch). Returns structured summary (25-40 lines) with citations suitable for…
prototype-generator
Generates self-contained HTML prototypes for hypothesis validation. Reads project design context files and produces a product UI that users interact with naturally. Context separation ensures prototypes reflect product vision. Invoked by recipe-validate for Usability risk validation.
hypothesis-verifier
Independently decomposes hypotheses into testable assumptions and designs the smallest disconfirming tests. Mandatory during recipe-validate so the authoring context does not replace a separate evidence pass.
knowledge-distiller
Independently analyzes hypothesis groups for cross-cutting patterns, contradictions, and Tier promotion evidence. Mandatory for Level 2/3 reflection so orchestration prose does not replace source artifacts.
prd-reviewer
Reviews every recipe-define PRD for governing-outcome integrity, evidence, downstream usability, and verifiability. Invoked as the mandatory independent review before user approval.
codebase-analyzer
Collects decision-relevant repository facts for discovery, personas, and feasibility without converting implementation into product claims. Invoked by recipe-discover, recipe-validate, and recipe-persona.