Jeden Skill in Manus ausführen
mit einem Klick

Jeden Skill in Manus mit einem Klick ausführen

harbor-oracle-verifier

Sterne0

Forks0

Aktualisiert24. Juni 2026 um 09:08

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

Installation

Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.

In Manus ausführen

Quelle

hu-qi

hu-qi/tb-hard-studio

GitHub-Repository öffnen Creator-Repositorys ansehen

Download

In Manus ausführen

Verwandte BerufeSOC

Basierend auf der SOC-Berufsklassifikation

Softwarequalitätssicherungsanalysten und -testerInformatik- und Mathematikberufe·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.

Mehr aus diesem Repository

gleiches Repository

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.