Sandbox
@workersio/skills

Test-finding skill pack for coding agents

This repository packages one testing skill, `wio`, that guides an agent through test discovery, strategy, implementation, workload generation, review, and suite health checks. It is meant to help builders find bugs that matter before release, not just increase coverage.

166 stars26 forksPythonUpdated 1mo ago
Who it's for

Builders who want their coding agent to find real regressions, choose stronger tests, and review test quality instead of writing shallow assertions.

What it delivers

You can use your agent to add tests that catch meaningful bugs and reject tests that do not protect the product.

What it does

Scan for test candidates

`$wio scan` maps product behavior, recent changes, existing tests, CI, and risk areas before choosing what to test.

Run the full test loop

`$wio test` guides candidate discovery, strategy selection, test writing, and review in one workflow.

Generate realistic workloads

`$wio workload` creates seeded, replayable sessions or traffic that expose bugs beyond simple wrappers and parameter sweeps.

Review test value

`$wio review` judges a test as `KEEP`, `REDO`, or `REMOVE` based on whether it catches a real regression.

Audit suite health

`$wio doctor` looks for weak assertions, over-mocking, flakes, skipped tests, slow loops, and CI blind spots.

Use focused subagents

The repo includes `wio-candidate-scout`, `wio-strategy-critic`, and `wio-test-reviewer` for read-only discovery, strategy critique, and post-write review.

How to get it

  1. 1Direct skill install
    npx skills add workersio/skills
  2. 2Codex plugin marketplace
    codex plugin marketplace add workersio/skills
  3. 3Claude Code plugin marketplace
    claude plugin marketplace add workersio/skills
    claude plugin install wio@workersio-skills
  4. 4For local Codex plugin testing from this checkout
    codex plugin marketplace add .

README

@workersio/skills running inside a coding agent terminal

Skills to find unknown bugs before release!

We help your agents test edge cases in your software where rare production bugs hide.

License Stars

Most AI-written tests optimize for coverage. They assert implementation details, mock away the real risk, and pass even when the product breaks. @workersio/skills gives coding agents a testing workflow that asks a stricter question: will this test catch a real regression that users, operators, reviewers, or maintainers care about?

@workersio/skills is one skill with five command modes:

$wio scan      # Find high-value tests to add next.
$wio test      # Run candidate discovery, strategy, implementation, and review.
$wio workload  # Generate realistic, replayable workloads with new bug-finding value.
$wio review    # Judge whether a test should be kept, redone, or removed.
$wio doctor    # Audit suite health, weak assertions, flakes, mocks, and CI blind spots.

Works With

CodexCodex    Claude CodeClaude Code    CursorCursor    GitHub CopilotGitHub Copilot    GeminiGemini

@workersio/skills is packaged for Codex and Claude Code today. The core skill content is plain Markdown, so the testing workflow can also be adapted by other coding agents that support project skills, instructions, or reusable prompts.

Install

Direct skill install:

npx skills add workersio/skills

Codex plugin marketplace:

codex plugin marketplace add workersio/skills

Claude Code plugin marketplace:

claude plugin marketplace add workersio/skills
claude plugin install wio@workersio-skills

For local Codex plugin testing from this checkout:

codex plugin marketplace add .

What It Provides

AreaWhat the agent gets
Test discoveryMaps product behavior, recent changes, existing tests, CI, and risk areas before choosing what to test.
Strategy selectionChooses the right level: unit, component, integration, contract, E2E, workload, fuzz, property, mutation, resilience, or monitoring.
Test writingAdds focused tests using the repository's own framework, naming, fixtures, helpers, and runner conventions.
Workload generationCreates realistic, adversarial, seeded, replayable sessions or traffic that add coverage beyond wrappers and parameter sweeps.
Test reviewApplies a strict value gate: KEEP, REDO, or REMOVE. Coverage alone is not enough.
Suite healthFinds weak assertions, over-mocking, broad snapshots, flakes, skipped tests, slow feedback loops, and CI blind spots.

Command Guide

CommandUse it whenExample
$wio scan [target]You do not yet know what to test.$wio scan checkout
$wio test [target]You want the full write-and-review loop.$wio test billing eligibility regression
$wio workload [target]The risk lives in realistic user, API, CLI, job, or load behavior.$wio workload onboarding session
$wio review [target]A test exists and you need to decide whether it has real value.$wio review tests/billing_eligibility_test.py
$wio doctor [target]The suite is hard to trust or maintain.$wio doctor API test suite

Quality Bar

An approved test should answer:

  • What user, operator, customer, or API consumer failure does this prevent?
  • What production, release, support, debugging, or review risk does it reduce?
  • Would it fail for the regression that matters?
  • Which assertion or invariant catches the plausible bug?
  • Does the setup preserve the important dependency, state, permission, timing, or data risk?
  • Does this belong in local development, PR CI, nightly, release, or production monitoring?
  • If this is a workload, what existing workload gap does it fill?

If those answers are weak, the test should be redesigned or removed.

Subagents

@workersio/skills includes three optional focused subagents:

SubagentRole
wio-candidate-scoutRead-only discovery of high-value test candidates before implementation.
wio-strategy-criticRead-only challenge of the selected strategy before editing tests.
wio-test-reviewerRead-only post-write review that returns KEEP, REDO, or REMOVE.

The main agent still writes the test and owns the final decision. Subagents inspect, challenge, and review; they do not duplicate the reference library or replace the main workflow.

Repository Layout

PathPurpose
plugins/wio/skills/wio/SKILL.mdSource of truth for the skill workflow.
plugins/wio/skills/wio/references/Detailed testing guidance loaded only when relevant.
plugins/wio/agents/Claude Code plugin subagents.
.codex/agents/Codex custom-agent TOML files for project or user installs.
plugins/wio/hooks/hooks.jsonShared plugin hook config.
.agents/plugins/marketplace.jsonCodex marketplace entry.
.claude-plugin/marketplace.jsonClaude Code marketplace entry.
public/README assets, demo GIF, and agent icons.

Codex Agents And Hooks

Codex plugin installs include the skill and plugin hook config. Codex custom agents are a separate native surface loaded from .codex/agents/ in a project or ~/.codex/agents/ for the user.

Enable the Codex agents globally:

mkdir -p ~/.codex/agents
cp .codex/agents/wio-*.toml ~/.codex/agents/

Enable the Codex agents for the current project:

mkdir -p .codex/agents
cp /path/to/wio-skills/.codex/agents/wio-*.toml .codex/agents/

Enable the Codex hook config for the current project:

mkdir -p .codex
cp /path/to/wio-skills/.codex/hooks.json .codex/hooks.json

Claude Code Agents And Hooks

Claude plugin installs include the skill, plugin hooks, and Markdown subagents from plugins/wio/agents/. Project-local Claude Code agents can also be copied into .claude/agents/ when a repository is not using the plugin.

References

Detailed testing guidance lives only in plugins/wio/skills/wio/references/. Reference topics cover behavior mapping, risk-based testing, test levels, workload modeling, oracles, fixtures, mocks, suite health, static analysis, security testing, fuzzing, property-based testing, mutation testing, performance testing, resilience testing, and regression selection.

Contributing

Keep the public surface area small: one skill, wio, with command modes scan, test, workload, review, and doctor.

Detailed testing guidance belongs in plugins/wio/skills/wio/references/, not duplicated inside cloud folders, subagents, hooks, or extra skill trees. When adding a reference topic, add both overview.md and tools.md, then link it from plugins/wio/skills/wio/references/index.md.

See CONTRIBUTING.md for contribution workflow and release expectations.

License

MIT. See LICENSE.

Files in the repo

Repository payload13 top-level entries
  • .agents
  • .claude
  • .claude-plugin
  • .codex
  • plugins
  • public
  • AGENTS.md
  • CHANGELOG.md
  • CLAUDE.md
  • CONTRIBUTING.md
  • LICENSE
  • README.md
  • SECURITY.md

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More skills

obra/
superpowers

An agentic skills framework & software development methodology that works.

285k
1 add

817 structured cybersecurity skills for AI agents · Mapped to 6 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND, NIST AI RMF & MITRE F3 (Fight Fraud) · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platforms · 29 security domains · Apache 2.0

33k
Da7-Tech/
SureForge

Agent Skill for complex work: research before asking, ask before planning, plan before building, verify before delivering, independent review before calling it done. Plain text, no runtime.

84
Leonxlnx/
unlazy

Anti-laziness skill for AI agents. Core: the Depth Tree method, which splits a task N layers deep and gives every leaf the full time budget of the whole task, so effort multiplies with depth. Grounded in 2025-2026 research on model laziness, underthinking and premature completion.

3.2k
Tiger3807861189/
J-Space-Cognition-Suite-V3.7

J-Space Cognition Suite V3.7 - AI cognitive-enhancement Skills based on Anthropic's J-space global workspace research. | 哔哩哔哩:Tiger380 (UID 3494375382321675) — https://space.bilibili.com/3494375382321675

3k