OpenSpec Spec Compress is a bundled preset for mechanical cleanup work inside openspec/specs/. It packages the documented compression workflow into a reusable preset-task entry instead of relying on hand-written one-off prompts.
Why use OpenSpec Spec Compress
It keeps the write boundary explicit
The preset is designed around openspec/specs/ and openspec/specs/archive/. It tells the executor to stay away from active openspec/changes/ content unless the user explicitly expands scope.
It standardizes the safe compression workflow
The prompt carries the same guardrails described in docs/openspec-spec-compression.md: establish a validation baseline, flatten delta headers, archive retired specs with git mv, preserve active requirements, and rerun validation incrementally.
It is automation-friendly by default
Non-interactive defaults are built into the prompt. The executor should not stop for optional clarification when a reasonable assumption keeps the workflow moving.
Best fit
- Best for: reducing structural OpenSpec spec debt without redesigning the capabilities themselves.
- Also good for: archive sweeps, delta-header cleanup, and mechanical spec normalization.
- Not for: semantic requirement rewriting, missing
SHALL / MUST fixes, missing Scenario authoring, or broad capability redesign.
OpenSpec Spec Compress 是一个面向 openspec/specs/ 机械清理工作的内置 preset。它把已经写进 docs/openspec-spec-compression.md 的压缩流程产品化,不再依赖一次性的手工 prompt。
为什么使用 OpenSpec Spec Compress
它把写入边界写死在 prompt 里
这个 preset 以 openspec/specs/ 与 openspec/specs/archive/ 为核心写入范围,并要求执行器默认不要触碰 active openspec/changes/,除非用户明确扩展范围。
它固化了安全的压缩节奏
提示词直接携带压缩指南里的关键规则:建立 validate 基线、扁平化 delta 头、用 git mv 归档退役 spec、保留活跃 requirement,并且每个逻辑步骤后都重新校验。
它默认适配自动化环境
非交互默认行为已经内置在 prompt 中。只要合理假设能让流程继续,执行器就不应停下来追问可选细节。
适用场景
- 最适合:在不重做 capability 设计的前提下,降低 OpenSpec spec 的结构性债务。
- 也适合:归档清扫、delta 头清理、机械式规范化处理。
- 不适合:语义级 requirement 改写、补写
SHALL / MUST、补写 Scenario,或大规模 capability 重设计。