github.devotd.
ReleaseCoding agents

code-yeongyu/oh-my-openagent v5.1.35 — Typecheck and sync fixes

code-yeongyu/oh-my-openagent v5.1.35 fixes type package references, memory sync merging, and categories configuration error guidance.

Repositorycode-yeongyu/oh-my-openagent ↗

OmO: Just type "mass ulw" keyword with your prompt. Now you are the master of graph engineering.

stars
★ 70.1K
Forks
5.8K
Language
TypeScript
License
NOASSERTION
Latest release
v5.1.35
Read original source ↗View on GitHub ↗

code-yeongyu/oh-my-openagent v5.1.35 is a release that changes how typechecking resolves package references, along with memory sync and categories configuration guidance. It was published on October 17, 2026, and the changes are recorded in the GitHub release. The repository description mentions a workflow of typing “mass ulw” into the prompt and graph engineering. Rather than rewriting that description, this version addresses inspection, sync, and configuration error guidance in the development process.

The typecheck pulled @types/node from the parent folder of the checked-out repository, which could make local and CI results differ even on the same commit. v5.1.35 declares its own @types/node and bun-types, and reads type packages only from its own node_modules. This is therefore a change that scopes the packages used for type checking to inside the repository. For memory sync, it now merges captures from stale devices as concurrent edits to prevent lost work. And categories entries that used agent names as-is are no longer silently ignored; a warning and a corrected configuration example are shown instead.

This change addresses the problem of diverging type check results, merging of stale captures, and guidance for certain categories settings. However, the provided release information contains no performance figures or benchmarks, so the check speed or the magnitude of the sync improvement cannot be known. The full scope of the settings covered by the categories warning was not given either. What can be confirmed is the fix to the explicitly stated error handling and guidance, and the basis for judging any effect beyond that is limited.