onsager-ai · Agent Skills

共 5 个 Skill，均有经来源核对的实质文档。打开卡片可阅读原始说明，了解包内文件和安装方式。

onsager-ai

社区

manage-lane

Maintain user intent and task continuity in the current Manager session for an existing engineering lane, including status, amendments and handover. Use orchestrate-harnesses for engineering planning and execution.

1小时安装量

2

GitHub Stars

0

最近更新

10月27日

MIT2 文件

查看详情 →

onsager-ai

社区

orchestrate-harnesses

Manage engineering work with native harness subagents as Workers, cost-conscious model selection, and HAPI synchronization between Managers across sessions and machines.

1小时安装量

2

GitHub Stars

0

最近更新

10月27日

MIT5 文件

查看详情 →

onsager-ai

社区

prompt-for-decisions

Present a genuinely unresolved user choice in the current Manager session with evidence, a recommendation and meaningful trade-offs. Use for material scope or authority boundaries, not routine engineering decisions owned by the Manager.

1小时安装量

2

GitHub Stars

0

最近更新

10月27日

MIT1 文件

查看详情 →

onsager-ai

社区

workbench-setup

Install, inspect, update, or remove Agent Workbench shared global instructions and its local launcher while preserving personal instruction text.

1小时安装量

2

GitHub Stars

0

最近更新

10月27日

MIT3 文件

查看详情 →

onsager-ai

社区

codegraph

Pre-indexed code knowledge graph (MCP, SQLite + tree-sitter) for faster, lower-token exploration of brownfield codebases. Use when starting work on a repo larger than ~500 files or when the task involves cross-file traversal — "where is X used", "what calls Y", "what breaks if I change Z", "trace flow from A to B", "explain this subsystem". Skip for single-file edits or sessions shorter than the cold-start cost. Triggers include "codegraph", "code graph", "index this repo", "where is X defined",

1小时安装量

1

GitHub Stars

5

最近更新

8月12日

MIT4 文件

查看详情 →