task-implementer-test

Verifies Task Implementer contracts and disposable multi-tier application lifecycles.

7|5|Updated Jul 22, 2025
One-click install
npx skills add https://github.com/nebius/nebius-ps-services --skill task-implementer-test-nebius
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: task-implementer-test
Source: https://github.com/nebius/nebius-ps-services/tree/main/skills/task-implementer-test
Command: npx skills add https://github.com/nebius/nebius-ps-services --skill task-implementer-test-nebius

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? It provides an explicit, isolated way to verify that the Task Implementer skill works correctly, running deterministic contract checks and optionally exercising a real disposable frontend/API/PostgreSQL fixture without touching user projects. ## Core Features & Use Cases - Deterministic Verification: Run the no-flag suite to validate Task Implementer metadata, public interface parity, workspace, scheduler, and Git wave contracts without Docker or real projects. - Disposable Live Lifecycle: Use --create to seed a local-only brownfield fixture, drive the real Task Implementer workflow, launch a labeled Docker Compose stack, and validate application and database semantics. - Owned Cleanup and Retention: Use --create --keep to retain one generation for inspection and --destroy to remove exactly that owned generation, with generation-fenced state and sanitized stage-matrix reports. - Use Case: After modifying Task Implementer, run the verifier to confirm worker isolation, digest-bound handoffs, fast-forward promotion, and an unchanged-prompt ALREADY_COMPLETE rerun before releasing. ## Quick Start Ask the agent to run $task-implementer-test to execute the deterministic verification suite and produce a sanitized report.

Frequently Asked Questions about task-implementer-test

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

FAQPage Schema
How do I verify the Task Implementer skill works correctly?

Invoke $task-implementer-test with no flags to run the deterministic contract, workspace, scheduler, and Git wave suites. It validates the public interface and helper behavior without Docker, workers, or real projects, then writes a sanitized report.

How do I test Task Implementer on a real application?

Run $task-implementer-test --create to seed a disposable local-only brownfield fixture and drive the real public workflow. It verifies dependency waves, worker evidence, a running frontend/API/PostgreSQL stack, and restart persistence before cleaning up.

Does task-implementer-test work with both Codex and Claude?

Yes, it supports Codex via $task-implementer-test and Claude Code via /task-implementer-test or the plugin form. The lifecycle binds the selected agent through its isolated native home, and resume or cleanup requires the recorded host.

Can I keep the test application running for inspection?

Yes, use --create --keep to retain the current owned generation, runtime, and private state for inspection. Later run --destroy to remove exactly that generation; repeating destroy returns ALREADY_DESTROYED.

When should I not use task-implementer-test?

Do not use it for ordinary implementation work or on user, production, customer, or remote-backed projects. It is an explicit verifier only and never replaces the real task-implementer workspace init, run, integrate, or remove commands.

Why does the verifier block with OWNERSHIP_BLOCKED?

Ownership checks fail on symlinks, wrong markers, remote Git repositories, external linked worktrees, generation mismatches, or ambiguous Docker resources. The verifier stops without cleanup or replacement so it never creates a second instance or deletes unowned state.