`loop-agent` 是面向 AI coding agent 的仓库级任务运行时和治理工具。它把一次研发任务组织成可生成、可校验、可执行、可恢复、可交接的 Agent DAG,并用 `.harness/`、`docs/` 和 shell verification 记录执行事实、长期治理资料和完成依据。
No confirmed malicious attack surface is established. The package is an explicit CLI automation/agent orchestrator with user-invoked shell, Pi, init, and local observer capabilities.
Package source references dynamic require/import behavior.
bin/loop-agent.jsView on unpkg · L16Package ships non-JavaScript build or shell helper files.
scripts/kb-graph-incremental-prepare.shView on unpkgThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/executors/shell-executor.jsView on unpkgPackage source references dynamic require/import behavior.
bin/loop-agent.jsView on unpkg · L16Package ships non-JavaScript build or shell helper files.
scripts/kb-graph-incremental-prepare.shView on unpkgThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/executors/shell-executor.jsView on unpkg