harn-testing

harn-testing is a skill for Claude Code from burin-labs/harn. It costs 24 tokens per session (1,871 once invoked), scanned A, original, Apache-2.0.

A guide for testing the part of a system that owns a behavior, at the smallest level that can disprove the claim. It covers unit tests, workflow tests, end-to-end tests, liveness, replay, and repeated trials for variable results.

In plain words
What is it for?
Use it to choose tests for parsers, runtime modules, workflows, adapters, product paths, progress and recovery, deterministic replay, or model quality.
Why use it?
It helps avoid tests that pass without checking the real interface or user path. It also makes test evidence and remaining uncertainty explicit.

Skill for Claude Code

Written for Claude Code: when-to-use in frontmatter.

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 skills/burin-labs/harn/harn-testing
Any agent
npx skills add burin-labs/harn --skill harn-testing
Clone the repo
git clone --depth 1 https://github.com/burin-labs/harn

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 harn-testing

README.md
[![agentmods](https://agentmods.dev/badge/skills/burin-labs/harn/harn-testing.svg)](https://agentmods.dev/skills/burin-labs/harn/harn-testing)
Your own site
<a href="https://agentmods.dev/skills/burin-labs/harn/harn-testing"><img src="https://agentmods.dev/badge/skills/burin-labs/harn/harn-testing.svg" alt="Measured on agentmods" height="20"></a>
Per session 24 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,871 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 5.1 $0.00024 $0.01871
Opus 5 $0.00012 $0.00936
Sonnet 5 $0.00005 $0.00374
Haiku 4.5 $0.00002 $0.00187

Measured 6d ago against content hash a532909eacd2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

harn-testing 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 6d 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.

crates/harn-skills/src/corpus/harn-testing/SKILL.md · 197 lines

How it starts

The opening of the file, as written. The whole thing — 197 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Harn testing

Use this skill to choose evidence that can falsify the claim being made.

Pair it with [[harn-probe]] for uncertain runtime facts, [[harn-orchestration]] for workflow ownership, and [[harn-product-quality]] for launch behavior.

Start from the claim

  • Write the claim in observable terms.
  • Name a plausible falsifier.
  • Identify the owning module and public interface.
  • Choose the lowest test altitude that can falsify the claim.
  • Add higher-altitude evidence when the claim crosses adapters or surfaces.
  • Do not count tests as proof; map evidence to claims.
  • Record what remains outside the evidence.

Test altitude

  • Pure language semantics: conformance fixture.
  • Parser, type, lint, or format behavior: focused crate test plus conformance.
  • Runtime module behavior: test through the public runtime interface.
  • Workflow behavior: deterministic harness test.
  • Adapter contract: production and in-memory adapters through the same seam.
  • Product behavior: canonical path end to end.
  • Liveness: progress, interruption, recovery, terminal state.
  • Model quality: multiple trials with a calibrated evaluator.

Determinism

  • Use mock_time, paused Tokio time, and explicit advancement.
  • Subscribe to events rather than polling.
  • Use OrchestratorHarness for workflow state.
  • Use deterministic provider and tool adapters.
  • Seed randomness and record the seed.
  • Avoid sleep, wall-clock deadlines, and short recv_timeout.
  • Avoid network dependencies in deterministic suites.
  • Assert on typed events and outcomes rather than incidental log order.

Interface tests

  • Tests and callers should cross the same interface.
  • Assert observable outcomes, errors, receipts, and state transitions.
  • Do not test through private seams merely because they are convenient.
  • Replace shallow-module tests when a deeper owning interface supersedes them.
  • Keep test setup smaller than the behavior under test.
  • Use production-shaped configuration defaults.
  • Verify closed error variants and diagnostic codes.
  • Keep fixtures minimal and readable.

Read the full file on GitHub · 197 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. 6d ago First seen · 197 lines · 24 tokens per session scan A a532909eacd2

Subscribe to this mod's changes

harn-testing is a skill published in the GitHub repository burin-labs/harn (20 stars, last pushed yesterday), licensed Apache-2.0. It adds 24 tokens to every session and 1,871 once invoked, about $0.0001 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 2026-08-30.