What is differential?

A comparison test that runs the project against pinned versions of python-docx and python-pptx, libraries used to create and read Word documents and PowerPoint presentations. It compares the resulting document structures across a set of test examples.

Use it to run document or presentation comparison tests for all or selected examples, using the available test harness and reporting missing presentation examples as skipped.

It detects differences between this implementation and the libraries it is intended to match. Pinned reference versions make changes in the reference library less likely to look like project regressions.

How tensorbee/rdocx configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to tensorbee/rdocx. 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/tensorbee/rdocx/main/.claude/commands/differential.md

Clone the repo

git clone --depth 1 https://github.com/tensorbee/rdocx

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.

README.md

[![agentmods](https://agentmods.dev/badge/commands/tensorbee/rdocx/differential/github.svg)](https://agentmods.dev/commands/tensorbee/rdocx/differential)

Your own site

<a href="https://agentmods.dev/commands/tensorbee/rdocx/differential"><img src="https://agentmods.dev/badge/commands/tensorbee/rdocx/differential/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.

Your own site · 80×15

<a href="https://agentmods.dev/commands/tensorbee/rdocx/differential"><img src="https://agentmods.dev/badge/commands/tensorbee/rdocx/differential.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 25 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,037 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.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00025	$0.01037
Opus 5.5	$0.00010	$0.00415
Sonnet 5.5	$0.00005	$0.00207
Haiku 4.5	$0.00003	$0.00104

Measured 26d ago against content hash fa5faa3435ec, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

differential 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 26d 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/commands/differential.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/differential [--oracle docx|pptx|all] [--fixture NAME]

Compare this workspace against the library it is a port of. Read .claude/skills/differential-testing.md first. It holds the rules, this holds the procedure.

Defaults to --oracle all.

Steps

Confirm the harness exists. If the differential harness has not shipped yet, say which story owns it, cite the backlog entry, and stop. Do not improvise a comparison inline. A one-off comparison that is not the harness produces a result nobody can reproduce.
Confirm the oracle is installed and pinned. Report the resolved version of python-docx or python-pptx before running anything. An unpinned oracle turns its upgrade into your regression, at a time you did not choose, so a drift between the pin and the resolved version stops the run.
Enumerate the corpus. Every fixture is constructed in code. There are no binary fixture files, per docs/hld/12-testing-strategy.md. The deck corpus is the one exception, it is fetched rather than committed, and a missing corpus is a skip to report rather than a failure to hide.
Run both sides per fixture, bounded by CPU count.
Compare the parsed tree, never the bytes. Attribute order, namespace prefix and whitespace are ours to decide, and docs/hld/04-opc-and-packaging.md sets that contract. A byte diff against a Python writer fails on decisions that were made deliberately.

For a render comparison, rasterise both sides at a stated DPI and compare with a stated metric and threshold. Deterministic font mode is mandatory on our side.

Report per fixture, with the disagreeing path in the tree:

Differential corpus, oracle python-pptx 1.0.2: 41/42 pass, 1 fail.

FAIL  inherited_title_colour
  slide.shapes[0].text_frame.paragraphs[0].runs[0].font.color.rgb
    python-pptx: 1F497D
    rpptx:       000000
  Resolution chain stopped at the layout. See
  docs/hld/07-inheritance-and-resolution.md, "Walking to the master".

Read the full file on GitHub · 103 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.

26d ago First seen · 103 lines · 25 tokens per session scan A fa5faa3435ec

Subscribe to this mod's changes

differential is a command published in the GitHub repository tensorbee/rdocx (44 stars, last pushed yesterday), licensed MIT. It adds 25 tokens to every session and 1,037 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-04.

Documents Testing

report

Obsolète (retirée en 9.0) : utilisez /qa:status --report.

TheBeardedBearSAS/claude-craft · 21 tokens

brief-compliance-check

Check a LaTeX coursework submission against the requirements in a supplied PDF assessment brief. Use when verifying format, required sections, word limits, or deliverables before submission. Not for general prose proofreading; use $proofread.

flonat/flonat-research · 45 tokens

plugin-commands

:::info This document covers all build, test, and smoke commands for eIsland native plugins. Windows plugins live under plugins/windows/; macOS helpers live under plugins/macos/. Each plugin has its own package.json with independent scripts. ::.

JNTMTMTM/eIsland · 0 tokens

add-entry

Command "add-entry" from gosha70/code-copilot-team, covering /add-entry, procedure and refuse to merge if.

gosha70/code-copilot-team · 0 tokens

Qa

All of QA in one command — builds and maintains the Excel QA workbook (strategy, every module's test cases, role matrix, data rules, findings, run history with when/where each case passed) for the whole application, chosen modules or one feature, derived from the code first and a URL second; runs the cases with a hard…

Evyasys-Software-Solutions/EvyaGovernance · 143 tokens

validate-glossary

A read-only checker for a glossary, which is a shared list defining terms and field names. It checks whether terms are unique and whether the glossary agrees with requirements, database designs, interfaces, and prototypes.

ChenyCHENYU/wl-skills-design · 19 tokens

rdocx: Command for Claude Code