harbor-framework

harbor-framework/harbor

Instructions file CodexOpenCode

Instructions for harbor-framework/harbor, covering claude.md - harbor framework, contributing, project overview, quick start commands and install.

harbor-framework/harbor

Instructions file

Instructions for harbor-framework/harbor, a project described as: Framework for evaluating and improving agents.

harbor-framework/harbor

MCP server Claude CodeCodexCursor +2

MCP server "runtimeproof", hosted remotely at runtime-mcp-server, as configured in harbor-framework/harbor.

harbor-framework/harbor

Skill Claude CodeCodex

Existing task skill that should remain after job-level skill injection.

harbor-framework/harbor

Skill Claude CodeCodex

Write the proof file for the Harbor runtime skill injection example.

harbor-framework/harbor

Skill Claude CodeCodex

Generate a greeting message and write it to a file.

harbor-framework/harbor

Skill Claude CodeCodex

Scaffold a new Harbor benchmark adapter by running harbor adapter init and then guide implementation using the Adapters Agent Guide as the authoritative spec.

harbor-framework/harbor

Skill Claude CodeCodex

Create a new Harbor task for evaluating agents. Use when the user wants to scaffold, build, or design a new task, benchmark problem, or eval. Guides through instruction writing, environment setup, verifier design (pytest vs Reward Kit vs custom), and solution scripting.

harbor-framework/harbor

Skill Claude CodeCodex

Use when working with Harbor's harbor exec CLI workflow: compiling files, directories, or globs into Harbor tasks; running map jobs; configuring artifacts and existence-only verification; using map-reduce; writing or reviewing ExecConfig YAML/JSON/TOML; or debugging command behavior, config validation, and job outputs.

harbor-framework/harbor

Skill Claude CodeCodex

Publish a Harbor task or dataset to the registry. Use when the user wants to upload, publish, or share tasks or datasets/benchmarks on the Harbor registry.

harbor-framework/harbor

Skill Claude CodeCodex

Write Harbor task verifiers using Reward Kit. Use when creating or editing a task's tests/ directory, adding grading criteria, setting up LLM/agent judges, or designing verifiers that produce a reward score.

harbor-framework/harbor

Skill Claude CodeCodex

Create or reuse Hugging Face dataset PRs for harborframework/parity-experiments and upload Harbor parity/oracle result folders efficiently with sparse checkout, raw git pushes, and Git LFS.

harbor-framework/terminal-bench-1

Instructions file GitHub Copilot

Instructions for harbor-framework/terminal-bench-1: If the pull request implements a task for the Terminal-Bench benchmark (creates files in the tasks/ directory), then please do the following review of the task.

harbor-framework/terminal-bench-1

Instructions file

Instructions for harbor-framework/terminal-bench-1, covering claude.md, architecture, core components, common development commands and running tasks.

harbor-framework/terminal-bench-science

Skill Claude CodeCodex

Convert a Harbor benchmark task from Harbor's shared verifier mode (default) to separate verifier mode. Use when the user asks to "convert this task to separate verifier", "make the verifier run in its own container", or asks about Harbor's separate-verifier environment for a specific task.

harbor-framework/terminal-bench-science

Skill Claude CodeCodex

Review a benchmark task PR — downloads all artifacts, analyzes against the rubric, launches harbor view, and enables interactive review.

harbor-framework/terminal-bench-science

Skill Claude CodeCodex

Propose new rubric criteria based on review findings — creates a branch and PR.

harbor-framework/terminal-bench-science

Instructions file CodexOpenCode

Instructions for harbor-framework/terminal-bench-science, covering terminal-bench-science, repo structure, key documentation, for agents reviewing task prs and template relationship.

harbor-framework/terminal-bench-science

Instructions file

Instructions for harbor-framework/terminal-bench-science, a project described as: Terminal-Bench-Science: Evaluating AI agents on research workflows across scientific domains.

harbor-framework/skills

Plugin Claude Code

Plugin marketplace listing 1 plugin: harbor-skills.

harbor-framework/skills

Instructions file CodexOpenCode

Instructions for harbor-framework/skills, covering harbor-skills — agent quick reference, skill format quick reference, harbor documentation map, conventions and common pitfalls.

harbor-framework/skills

Instructions file

Instructions for harbor-framework/skills, a project described as: Public agent skills catalog for Harbor.

harbor-framework/skills

Skill Claude CodeCodex

Create Harbor benchmark adapters that convert external benchmark datasets into Harbor task format. Use when porting an existing benchmark to Harbor, running parity experiments, registering a dataset to the Harbor registry, or debugging adapter validation failures. Covers: adapter class interface (generatetask…

harbor-framework/skills

Skill Claude CodeCodex

Harbor CLI command reference and usage patterns. Covers harbor run, harbor jobs, harbor trials, harbor datasets, harbor adapters, harbor tasks, harbor view, harbor sweeps, harbor traces, harbor cache, and harbor admin commands. Use this skill whenever running Harbor evaluations, managing datasets, viewing results…