skills.rest
krzemienski
krzemienskiCOMMUNITY
Explore all skills by this author →
Website𝕏@nkrzemienskiGitHub

forge-execute

Execute end-to-end validation journeys with per-attempt evidence and automated fix attempts.

2|Updated Apr 8, 2026
Search Tags:#e2e#validation#evidence#autofix#state-management#ci#rebuild
One-click install
npx skills add https://github.com/krzemienski/validationforge --skill forge-execute-krzemienski
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: forge-execute
Source: https://github.com/krzemienski/validationforge/tree/main/skill-audit-workspace/forge-execute/skill-snapshot
Command: npx skills add https://github.com/krzemienski/validationforge --skill forge-execute-krzemienski

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Provides a deterministic, real-system validation runner that removes reliance on mocks and fragile unit-only checks by executing end-to-end journeys against the actual application, capturing verifiable evidence, and iteratively repairing failures until stability or a hard abort.

Core Features & Use Cases

  • Autonomous Forge Loop: Orchestrates PLAN → PREFLIGHT → EXECUTE → ANALYZE → FIX → REBUILD → RE-EXECUTE with a three-strike limit per journey.
  • Per-Attempt Evidence: Creates isolated evidence directories for each attempt (e2e-evidence/forge-attempt-{N}/{journey-slug}/) and records outcomes and fixes in persistent state.
  • Safe Rebuild & Retry Rules: Enforces a required build step before any re-validation, supports full/quick/ci/targeted modes, and emits CI-friendly exit codes.
  • State Persistence & Reporting: Persists run and journey state to .validationforge/forge-state.json and generates a unified report at e2e-evidence/report.md.
  • Use Case: Run in CI to detect regressions that unit tests miss, or run locally to validate and iteratively fix a failing login or settings journey while preserving evidence.

Quick Start

Start the autonomous validation loop by running forge-execute after creating a validation plan so it captures per-attempt evidence, attempts fixes up to three strikes, and produces a unified report.

Frequently Asked Questions about forge-execute

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run end-to-end validation that automatically fixes failures?▼

End-to-end validation with autonomous fix-and-retry executes journeys against the real system, captures per-attempt evidence, and iteratively repairs failures up to a three-strike limit before aborting.

What is the best way to validate CI pipelines without relying on mocks?▼

Validating CI pipelines without mocks requires a deterministic runner that executes end-to-end journeys against the actual application, captures verifiable evidence, and emits CI-friendly exit codes.

How does automated fix-and-retry work for failing checkout or login flows?▼

Automated fix-and-retry orchestrates a loop of execute, analyze, fix, and rebuild steps for failing journeys, maintaining isolated evidence directories per attempt and persisting state to a forge-state.json file.

Do I need to rebuild my application before re-running e2e validation?▼

Yes, a required build step is enforced before any re-validation to ensure changes are applied, supporting full, quick, CI, and targeted execution modes for the retry loop.

How is validation evidence stored for repeated e2e test attempts?▼

Validation evidence is stored in isolated directories per attempt under e2e-evidence/forge-attempt-{N}/{journey-slug}/, with run and journey state persisted to forge-state.json and a unified report generated at report.md.

Related Skills

Explore More Software Engineering Skills→
krzemienskikrzemienski
2

create-validation-plan

Create validation plans with PASS criteria and ordered user journeys.

Community
Intermediate
krzemienskikrzemienski
2

retrospective-validation

Analyze historical validation runs and deployments to compute false pass/fail rates and confidence scores.

Community
Intermediate
notquenotque
415

endpoint-validator

Validate API endpoint responses, timing, and JSON keys from endpoints.json.

Community
Intermediate
ricablericable
1

Claude Flow Testing

Run TDD tests for Claude Flow V3 modules with mocks and fixtures.

Community
Intermediate