Ejecuta cualquier Skill en Manus
con un clic

Ejecuta cualquier Skill en Manus con un clic

harbor-oracle-verifier

Estrellas0

Forks0

Actualizado24 de junio de 2026 a las 09:08

Implement a legitimate Harbor Oracle solution and validate its behavior against an independent verifier without weakening the task contract.

Instalación

Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.

Ejecutar en Manus

Fuente

hu-qi

hu-qi/tb-hard-studio

Abrir repositorio de GitHub Ver repositorios del creador

Descarga

Ejecutar en Manus

Ocupaciones relacionadasSOC

Basado en la clasificación ocupacional SOC

Analistas de garantía de calidad de software y probadoresOcupaciones informáticas y matemáticas·SOC 15-1253

SKILL.md

readonly

name	harbor-oracle-verifier
description	Implement a legitimate Harbor Oracle solution and validate its behavior against an independent verifier without weakening the task contract.
argument-hint	<case_id>
user-invocable	true
context	fork
agent	oracle-solution-engineer
model	opus
effort	xhigh
allowed-tools	Read Write Edit Glob Grep Bash

Harbor Oracle Verification

Target case: $ARGUMENTS

The Oracle proves solvability. Work only in cases/<case_id>/task/solution/ unless the main agent explicitly assigns a verifier defect investigation. Do not weaken or redesign tests, public constraints, or task semantics to make the Oracle pass.

Solve from the same baseline available to a competent solver. Do not hardcode public samples or read private briefs, calibration evidence, hidden grader material, or authoring assets. Run make oracle CASE=<case_id> from a clean environment and preserve command/log metadata under evidence/<case_id>/oracle/.

Return changed files, run command, reward, logs, and any independently reviewable coverage concern.

Más de este repositorio

mismo repositorio

harbor-environment-design

hu-qi/tb-hard-studio

Build or repair a deterministic Harbor task environment with explicit resources, network policy, health checks, and safe solver-visible boundaries.

2026-06-240

harbor-redteam-verifier

hu-qi/tb-hard-studio

Red-team a Harbor verifier for answer leakage, reward hacking, shared-container tampering, artifact manipulation, and semantic under-testing.

2026-06-240

harbor-rollout-calibration

hu-qi/tb-hard-studio

Calibrate a frozen tb-hard task using Harbor Oracle and real-agent rollouts, then distinguish authentic difficulty from ambiguity, flakiness, leakage, or infrastructure failure.

2026-06-240

harbor-task-create

hu-qi/tb-hard-studio

Create or evolve a canonical Harbor task inside a stable tb-hard case directory. Use after the design brief names observable acceptance criteria and at least three difficulty dimensions.

2026-06-240

tb-hard-release-audit

hu-qi/tb-hard-studio

Audit a calibrated Harbor task for deterministic purchaser export, leakage, evidence completeness, and release readiness.

2026-06-240

tb2-pattern-research

hu-qi/tb-hard-studio

Research Harbor and Terminal-Bench 2 patterns to design an original tb-hard task. Use before approving a design brief or when checking whether a candidate is too shallow or derivative.

2026-06-240

name	harbor-oracle-verifier
description	Implement a legitimate Harbor Oracle solution and validate its behavior against an independent verifier without weakening the task contract.
argument-hint	<case_id>
user-invocable	true
context	fork
agent	oracle-solution-engineer
model	opus
effort	xhigh
allowed-tools	Read Write Edit Glob Grep Bash

Harbor Oracle Verification

Target case: $ARGUMENTS

Return changed files, run command, reward, logs, and any independently reviewable coverage concern.