DeepReason: Skill for Claude Code

.claude/skills/dr-reproduce/SKILL.md

dr-reproduce is a skill for Claude Code from AHepi/DeepReason. It costs 42 tokens per session (652 once invoked), scanned A, original, MIT.

A workflow for proving the cause of a diagnosed software defect with a small offline test or script. It runs against existing records or minimal in-memory state instead of starting a live provider run.

In plain words
What is it for?
Use it after a diagnosis to create REPRO.md and a runnable artifact showing the defect now and checking that it disappears after a future fix.
Why use it?
It turns a diagnosis into repeatable evidence before anyone changes production code. Offline reproduction is cheaper and more predictable than relying on a live service.

Skill for Claude Code

Written for Claude Code: installed under .claude/.

This is AHepi/DeepReason's own configuration. It tells Claude Code how to work on DeepReason itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything DeepReason configures →

Reuse

Borrowing it

Nothing to install: this file belongs to AHepi/DeepReason. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/AHepi/DeepReason/main/.claude/skills/dr-reproduce/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/AHepi/DeepReason

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 dr-reproduce

README.md
[![agentmods](https://agentmods.dev/badge/skills/ahepi/deepreason/dr-reproduce/github.svg)](https://agentmods.dev/skills/ahepi/deepreason/dr-reproduce)
Your own site
<a href="https://agentmods.dev/skills/ahepi/deepreason/dr-reproduce"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-reproduce/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for dr-reproduce

Your own site · 80×15
<a href="https://agentmods.dev/skills/ahepi/deepreason/dr-reproduce"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-reproduce.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 42 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 652 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 $0.00042 $0.00652
Opus 5 $0.00021 $0.00326
Sonnet 5 $0.00008 $0.00130
Haiku 4.5 $0.00004 $0.00065

Measured 9d ago against content hash 4bfe20268ba3, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

dr-reproduce 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 9d 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.

.claude/skills/dr-reproduce/SKILL.md · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Reproduce the cause

Input: DIAGNOSIS.md's falsifiable prediction. Output: REPRO.md plus one runnable artifact that shows the defect NOW (and will show its absence after the fix). You still change no production code.

Choose the cheapest sufficient form, in this order

  1. Record replay — the defect is already capted in a committed run root: a script that runs verify_root(<root>) or walks the log and prints the violating fact. Zero live cost; deterministic.
  2. Offline unit reproduction — construct the minimal state in a test: register the same problems/artifacts/policies the record shows and assert the wrong behavior happens. Reuse existing test helpers (_prepare_run, controller fixtures, _policy builders) from the nearest tests/test_*.py; do not invent new scaffolding when a helper exists.
  3. Minimal in-memory check — a 20-line python3 heredoc proving the mechanism (e.g. round-trip a receipt through canonical_json and show key order changes). Acceptable as evidence, but pair it with form 1 or 2 for the regression artifact.

NEVER reproduce by launching a live provider run. Live runs are for dr-verify-outcome, at most once, and only if the goal demands it.

Fidelity rules

  • The reproduction must mirror the live conditions the record shows, not a convenient simplification. If admission auto-accepted import-role artifacts before cycle 0, your fixture registers those artifacts too. A reproduction that passes for a different reason than the live failure will approve a wrong fix.
  • One assertion states the DEFECT (currently failing or currently printing the violation), phrased so it inverts cleanly post-fix.
  • Respect frozen-record invariants in fixtures: one manifest sha per capability chain, constant fence seqs within a proposal's chain — fixture WellFormednessError means your fixture is wrong, not the harness.

REPRO.md template

# Reproduction
Form: record-replay | unit-test | in-memory
Artifact: <path (test id) or inline command>
Current output: <paste the failing/violating output, trimmed>
Confirms diagnosis: yes — <one line linking output to mechanism>
Post-fix expectation: <exact output/assertion after a correct fix>

Read the full file on GitHub · 64 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. 9d ago First seen · 64 lines · 42 tokens per session scan A 4bfe20268ba3

Subscribe to this mod's changes

dr-reproduce is a skill published in the GitHub repository AHepi/DeepReason (142 stars, last pushed 2d ago), licensed MIT. It adds 42 tokens to every session and 652 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 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

4-phase root cause debugging: understand bugs before fixing.

NousResearch/hermes-agent · 16 tokens

langsmith-observability

LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.

davila7/claude-code-templates · 45 tokens

experimental-code-coverage-local-debugger

Runs code coverage locally via Universal Test Runner (UTR) or helper scripts, mimicking LUCI trybots. Activate when CQ tryjobs fail or underreport coverage, to test local GN/recipe repairs before uploading, or to debug hermetic crashes.

chromium/chromium · 59 tokens

adversarial-reviewer

Adversarial code review that assumes bugs exist and hunts for them. Use when asked to review code, find bugs, audit for correctness, stress-test a PR, or when someone says "tear this apart" or "what's wrong with this". Give no benefit of the doubt — every line is guilty until proven innocent.

emdash-cms/emdash · 71 tokens

cli-e2e

Write, modify, or debug Docker-based Composio CLI end-to-end tests under ts/e2e-tests/cli, including binary invocation, fixture isolation, output assertions, and package manifests. Use for CLI E2E test suites only; use cli-command for CLI source implementation.

ComposioHQ/composio · 62 tokens

ios-simulator

Verify and debug native, React Native, Expo, or Flutter apps on an iOS Simulator with agent-device. Use when an agent needs to launch an app, inspect its live UI, tap, type, scroll, validate a code change, collect failure evidence, or reproduce a workflow on an iPhone or iPad Simulator.

callstack/agent-device · 69 tokens