# Validate

> Quality gate — testing, linting, validating. Use after implement. Runs tests and checks quality gates.

- **Type:** Skill
- **Install:** `agentstack add skill-bugroger-beastmode-validate`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [BugRoger](https://agentstack.voostack.com/s/bugroger)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [BugRoger](https://github.com/BugRoger)
- **Source:** https://github.com/BugRoger/beastmode/tree/main/plugin/skills/validate

## Install

```sh
agentstack add skill-bugroger-beastmode-validate
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# /validate

Verify code changes meet quality standards before release.

No release without passing validation.

## Guiding Principles

- All features must complete implementation before validation begins
- Fix failures before reporting — attempt repairs, don't just stop at first error
- Gates are pass/fail — no partial credit, no "close enough"
- Validate the artifact, not the process — results matter, not how we got there

## Phase 0: Pre-Execute

### 1. Check Feature Completion

Scan for implementation artifacts to verify all features have been implemented:

```bash
ls .beastmode/artifacts/implement/*-$epic-*.md 2>/dev/null
```

Cross-reference against the feature plan files to determine completion status.

Print status:

```
Feature Completion Check
────────────────────────
✓ feature-1 — completed
✓ feature-2 — completed
✗ feature-3 — pending

Result: BLOCKED — 1 feature still pending
```

If any features are NOT completed:
- Print which features are pending
- STOP — do not proceed to test execution
- Suggest: "Run `/beastmode:implement -` to complete remaining features."

If all completed: proceed to next step.

### 2. Identify Test Strategy

From context, determine:
- Test command (e.g., `npm test`, `pytest`)
- Lint command (if configured)
- Type check command (if configured)
- Custom gates from design acceptance criteria

## Phase 1: Execute

### 1. Run Tests

```bash

```

Capture output and exit code.

### 2. Run Lint (if configured)

```bash

```

### 3. Run Type Check (if configured)

```bash

```

### 4. Run Custom Gates

Execute any custom gates defined in `.beastmode/context/VALIDATE.md`.

## Phase 2: Validate

### 1. Analyze Results

Check each gate:
- Tests: PASS/FAIL
- Lint: PASS/FAIL/SKIP
- Types: PASS/FAIL/SKIP
- Custom: PASS/FAIL/SKIP

### 2. Determine Overall Status

- All required gates pass → PASS
- Any required gate fails → FAIL

### 2b. Identify Failing Features

If any test gate fails, identify which features are responsible:

1. Parse test output for feature-scoped test names (e.g., `*.integration.test.ts`, tagged with `@`)
2. Map failures to feature slugs using naming conventions
3. Record per-feature pass/fail results

If feature-level identification is not possible (tests are not feature-scoped), fall back to blanket failure.

### 3. Generate Report

```markdown
# Validation Report

## Status: {PASS|FAIL}

### Tests
{output}

### Lint
{output or "Skipped"}

### Types
{output or "Skipped"}

### Custom Gates
{output or "None configured"}
```

## Phase 3: Checkpoint

### 1. Save Report

Save to `.beastmode/artifacts/validate/YYYY-MM-DD-.md` where `` is the epic name.

The validation report must begin with YAML frontmatter:

```
---
phase: validate
epic-id: 
epic-slug: 
status: passed
failed-features: feat-a,feat-b
---
```

Set `status` to `passed` or `failed` matching the validation result.

- `failed-features` is a comma-separated list of feature slugs that failed validation
- Only present when `status: failed` and specific features can be identified
- When absent on failure, the pipeline falls back to blanket regression (all features reset)

### 2. Commit and Handoff

If FAIL:
```
Validation failed.
Failing features: 
Re-dispatch count will increment for each failing feature (max 2 per feature).
The pipeline will automatically re-implement failing features.
```
STOP — do not proceed to commit.

If PASS:

Commit all work to the feature branch:

```bash
git add -A
git commit -m "validate(): checkpoint"
```

Print:

```
Next: beastmode release 
```

STOP. No additional output.

## Constraints

- No release without passing validation
- Do not proceed to commit if validation fails
- Do not proceed past Pre-Execute if features are incomplete

## Reference

### Quality Gates

**Why Gates Matter**

Quality gates prevent broken code from reaching release. Each gate is a checkpoint that must pass.

**Default Gates**

1. **Tests** (required) - All tests must pass
2. **Lint** (optional) - Code style compliance
3. **Types** (optional) - Type checking passes

**Custom Gates**

Add custom gates in `.beastmode/context/VALIDATE.md`:

```markdown
## Custom Gates

- [ ] Performance benchmarks pass
- [ ] Security scan clean
- [ ] Coverage > 80%
```

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [BugRoger](https://github.com/BugRoger)
- **Source:** [BugRoger/beastmode](https://github.com/BugRoger/beastmode)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-bugroger-beastmode-validate
- Seller: https://agentstack.voostack.com/s/bugroger
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.