shift-left-verification
Universal progressive testing and shift-left verification standard (Google Small Tests, Stripe level).
Skill by hemanth17r
Chalk Labs
Need help deploying shift-left-verification?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/hemanth17r-shift-left-verification
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: testing, writing
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/shift-left-verification
What's inside
- Skill name
shift-left-verification- Description
- Universal progressive testing and shift-left verification standard (Google Small Tests, Stripe level). Codifies fast deterministic unit tests (<100ms, zero I/O), atomic co-development (commit logic and test in the same changeset), milestone verification gates, and eliminates the 3 testing traps (mocking illusions, testing private implementation details, and pixel assertions). Trigger when writing unit tests, setting up test suites, verifying milestones, practicing TDD/co-development, designing testable architecture, or when asked to "write tests", "add tests", "test this", "need tests", "add test coverage", or "how should we test this".
- License
- MIT
- Instructions
- 85 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Shift-Left Progressive Verification Standard
## 1. The Core Philosophy: Verification as an Architecture Forcing Function
In high-performance engineering, testing is not an afterthought phase; it is an **architectural forcing function** executed continuously at each milestone.
* Code that is difficult to unit test is **badly architected code** (spaghetti coupling, hidden side-effects, hardcoded clocks, or direct network calls inside domain logic).
* Designing for testability naturally forces code to be **pure, modular, and dependency-injected**.
* Catching a bug at the local authoring gate takes **3 seconds**. Catching it in a staging or production environment takes **hours**.
---
## 2. Principle 1: Google's "Small Tests" Rule (The Zero I/O Mandate)
Classify unit tests strictly by execution physics. **80%+ of a project's tests must be Small Tests**:
```
┌────────────────────────────────────────┐
│ GOOGLE SMALL TEST STANDARD │
├────────────────────────────────────────┤
│ • Execution Time: < 100 milliseconds │
│ • Network Access: ZERO (Banned) │
│ • Disk File I/O: ZERO (Banned) │
│ • Sleep / Delays: ZERO (Banned) │
│ • Environment: Single Process │
└────────────────────────────────────────┘
```
* **Implementation**: Keep business calculations, state machines, math, data mappers, and validation logic pure and headless.
* **Execution**: A suite of 50 pure unit tests must execute in **under 100ms** locally (e.g. via `Vitest`, `Jest`, `pytest`), giving instant developer feedback with zero latency.
---
## 3. Principle 2: Atomic Co-Development & Micro-Commits
Never accumulate large unverified feature branches. Break work into atomic, milestone-gated slices:
1. **Write an Atomic Slice**: Implement a single pure unit of logic (30 to 100 lines).
2. **Co-Locate the Test**: Keep the test file immediately adjacent to the source:
```
src/features/bounties/
├── calculateReward.ts
└── calculateReward.test.ts
```
3. **Run the Verification Gate**: Confirm tests pass locally before touching downstream presentation or storage layers.
4. **Atomic Commit**: Commit the implementation and its corresponding test together in the **exact same commit**:
```bash
git commit -m "feat(bounties): implement dynamic sector reward calculation with unit test"
```
---
## 4. The 3 Testing Traps to Eliminate
Tests must accelerate development velocity, not become brittle maintenance liabilities.
### Trap 1: The "Mocking Illusion"
* ❌ **Anti-Pattern**: Mocking the database, the ORM, the network client, and external APIs. You end up testing mock configuration while production crashes on real schemas.
* ✅ **The Fix**: Never mock what you can keep pure. Test core logic with real in-memory inputs and outputs. Reserve real I/O verification for separate integration tests.
### Trap 2: Testing Implementation Details Instead of Contracts
* ❌ **Anti-Pattern**: Using spies to assert that an internal private method was invoked:
```ts
expect(service._privateCacheHelper).toHaveBeenCalledTimes(1);
```
Refactoring internal implementation breaks the test, even when the external behavior is 100% correct.
* ✅ **The Fix**: Treat components and functions as black boxes. **Assert public inputs and observable outputs/state changes only.**
### Trap 3: Unit-Testing Pixels & Styling
* ❌ **Anti-Pattern**: Writing unit tests asserting CSS classes or DOM heights (`expect(el).toHaveClass('px-4 py-2')`).
* ✅ **The Fix**: Test **state, data flow, and user interaction**, not visual styling. Verify that clicking an item toggles selection; verify visual aesthetics through browser inspection.
---
## 5. Pre-Flight Verification Checklist
Before marking any milestone or task complete, verify:
- [ ] **Zero I/O**: Do unit tests execute without network, file, or database roundtrips?
- [ ] **Sub-Second Speed**: Does the test suite run in milliseconds?
- [ ] **Co-Located Test Files**: Is the test placed directly next to the module it tests?
- [ ] **Contract-Only Assertions**: Are tests asserting public outputs rather than private implementation details?
- [ ] **Zero Pixel Tests**: Are tests free from fragile CSS class or DOM geometry assertions?
- [ ] **Edge Cases Covered**: Are boundary conditions (empty lists, null values, negative counts, zero thresholds) explicitly tested?
- [ ] **Atomic Commit Ready**: Is the unit test staged and committed alongside the feature code?
---
## 6. Progressive Slice-by-Slice Wiring & Responsiveness Gate
Never batch-dump 4-5 buttons, form fields, or interactive controls and test only after the entire feature is added. Every interactive control must be verified progressively as it is wired:
1. **Zero Dummy / Placeholder Controls**: Every button, input, or trigger must execute its real downstream handler, state transition, or persistence contract immediately upon being placed. Leaving unverified or dummy actions is strictly considered a bug.
2. **Tactile Responsiveness & Zero Shutter**: Controls must react instantly to touch/click (`active:scale-95`, loading spinners during async calls, zero UI freezing or event shutter).
3. **Headless State & Persistence Verification**: Verify that input data actually saves (localStorage, database, or state store) slice-by-slice before proceeding to stack subsequent UI controls.
4. **Interactive State Reversibility**: Verify toggle controls (open -> close -> open) and outside-click collision guards before declaring completion.From hemanth17r/ahr-workflow, MIT.
More from hemanth17r/ahr-workflow
- 0.645score
ground-up-engineering-protocol
hemanth17r
Universal architectural standard for autonomous coding agents to build any software system from scratch.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
grounded-founder-outreach
hemanth17r
Generates personalized, grounded, high-conviction founder outreach messages and application notes across all platforms (Wellfound, Y Combinator Work at a…
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
autonomous-division-of-labor
hemanth17r
Enforces maximum agent ownership and strict division of labor for setups, installations, and multi-device workflows.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
adversarial-mechanism-design
hemanth17r
Universal mechanism design, reverse game theory, incentive engineering, and adversarial bad-actor defense standard for ANY product, platform, or marketplace.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-db-resilience
hemanth17r
Universal database concurrency, production lock safety, and operational resilience standard.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-product-ux
hemanth17r
Universal product design, cognitive psychology, microcopy economy, and UX architecture standard (Apple, Google, Stripe level).
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-ui-craft
hemanth17r
Universal cognitive psychology and craftsmanship standard for high-conviction UI (Apple, Google, Linear, Stripe level).
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
blackbox-logic-flow-audit
hemanth17r
Universal programmatic black-box logic, wiring, and state flow verification standard across ANY software project.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-db-craft
hemanth17r
Universal database architecture and performance standard adapted from Big Tech for lean systems.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
feynman-intuitive-explainer
hemanth17r
Translates complex technical jargon, engineering architecture, and mathematical concepts into crystal-clear, intuitive real-world analogies (The Feynman…
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
fluid-tactile-motion
hemanth17r
Universal spring motion physics, tactile micro-interactions, and optimistic UI architecture standard (Apple iOS, Linear, LinkedIn level).
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
idea-reality-check
hemanth17r
Conducts a ruthless prior-art investigation, package shortcut check, and competitive reality check on technical tools, packages, or architectural proposals…
SkillAFreshness: A★ 22Claude Code +9
Signals
- Stars
- 22
- Forks
- 4
- Open issues
- 0
- Last commit (this folder)
- Oct 19, 2026 (15 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/hemanth17r/ahr-workflow
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from hemanth17r
- 0.645score
adversarial-mechanism-design
hemanth17r
Universal mechanism design, reverse game theory, incentive engineering, and adversarial bad-actor defense standard for ANY product, platform, or marketplace.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-product-ux
hemanth17r
Universal product design, cognitive psychology, microcopy economy, and UX architecture standard (Apple, Google, Stripe level).
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
autonomous-division-of-labor
hemanth17r
Enforces maximum agent ownership and strict division of labor for setups, installations, and multi-device workflows.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-db-craft
hemanth17r
Universal database architecture and performance standard adapted from Big Tech for lean systems.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-db-resilience
hemanth17r
Universal database concurrency, production lock safety, and operational resilience standard.
SkillAFreshness: A★ 22Claude Code +9 - 0.645score
big-tech-ui-craft
hemanth17r
Universal cognitive psychology and craftsmanship standard for high-conviction UI (Apple, Google, Linear, Stripe level).
SkillAFreshness: A★ 22Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get shift-left-verification running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
shift-left-verification skill. Agent Mart. https://agentmart.dev/skills/hemanth17r-shift-left-verification (last updated 2026-10-24)