Home · Reviews

All skill reviews

Every skill below was run on a standard task with and without it loaded, scored on our published 100-point rubric, and screened against six security red lines. Sorted by score — the evidence is on each page.

91/100
S

xlsx

Anthropic's spreadsheet skill: live formulas only, a professional formatting standard, and a mandatory LibreOffice recalculation pass before any workbook ships.

Official · AnthropicDeep testData AnalysisSecurity screen passed

“The first skill in our tests whose edge over the baseline is binary rather than a matter of taste: identical formula logic on both sides, but only the with-skill workbook shipped values a reader could actually see, because the skill mandates the recalculation pass nobody runs unprompted. Caveats: it needs LibreOffice installed, and the licence is proprietary.”

Read the full review →
88/100
A

writing-plans

The Superpowers planning skill: turns a spec into a bite-sized, TDD-shaped implementation plan — exact files, exact signatures, a failing test before every implementation step.

Deep testDevelopment WorkflowSecurity screen passed

“The strongest workflow skill we have tested. Its plan is executable by a developer who has never seen the codebase — 26 checkbox steps, 14 tests written as code, exact signatures in every task — where a competent baseline outline still leaves every interface to be invented. Docked for a description with no off-switch and a plan header that hard-wires Superpowers' own execution skills.”

Read the full review →
84/100
A

pdf

Anthropic's PDF cookbook: working recipes for extraction, merging, splitting, watermarks, encryption, OCR and form filling, plus eight helper scripts.

Official · AnthropicDeep testDocuments & OfficeSecurity screen passed

“On clean PDFs it ties with doing it yourself — the score comes from everything around that: twenty recipes that run as written, a proper form-filling validation loop, and the best file layering we have tested. Two honest caveats: it is proprietary-licensed, and it triggers on any PDF mention.”

Read the full review →
84/100
A

semgrep

Trail of Bits' approval-gated Semgrep scan: language detection, ruleset selection, a hard plan-approval step, parallel scans and merged SARIF — with telemetry off by construction.

Deep testSecuritySecurity screen passed

“The strongest process we have scored — hard approval gate, failure accounting, telemetry off by construction — wrapped around a scanner that missed the SQL injection in a 71-line file. Trust it to run a scan honestly; do not trust the scan alone to clear a file.”

Read the full review →
80/100
A

frontend-design

Anthropic's official skill for distinctive, non-templated visual design — aesthetic direction, typography and layout guidance for new UI.

Official · AnthropicDeep testFrontend & DesignSecurity screen passed

“The rare guidance skill whose live output is visibly less generic than baseline — docked mainly for a vague trigger description and zero worked examples.”

Read the full review →
77/100
B

github-actions-efficiency

GitHub's CI cost-cutting playbook from the awesome-copilot collection: audit a workflow's waste in a fixed order, apply guarded fixes — caching, stale-run cancellation, trigger scoping — and report savings without inventing them.

Official · GitHubDeep testDevOps & InfraSecurity screen passed

“A disciplined audit playbook whose guardrails and honesty rules are the best we have scored — held back by a description that hides its creation mode, a cache pattern one idiom behind the baseline, and a test workflow that differed from no-skill output in exactly one block: it cancelled stale runs.”

Read the full review →
73/100
B

webapp-testing

Anthropic's Playwright toolkit for testing local web apps: a reconnaissance-then-action pattern, a server-lifecycle helper script, and three worked examples.

Official · AnthropicDeep testTesting & QASecurity screen passed

“A sound browser-driving method wrapped around one useful script — reconnaissance before action, selectors from the rendered page, servers managed for you. It teaches an agent to operate a web app, not to judge one: there is no assertion guidance anywhere in its 95 lines, and its own server helper caused the only crash in our test run.”

Read the full review →
66/100
B

brand-guidelines

Anthropic's official brand card: seven exact colours and two font stacks for styling artifacts with Anthropic's look-and-feel.

Official · AnthropicDeep testMarketing & ContentSecurity screen passed

“Seventy-three lines of exact, copy-paste brand data that measurably beat a strong baseline on fidelity — 7 of 7 colours exact, both fonts right. But it is a data card, not a workflow: no steps, no checks, a trigger that can fire on other companies' brands, and feature text describing a PowerPoint script that does not ship.”

Read the full review →

Looking for a task first? Browse skills by category — or read how the scoring works.