github code-yeongyu/oh-my-opencode v3.0.0-beta.14

latest releases: v3.0.0-beta.16, v3.0.0-beta.15
10 hours ago

v3.0.0-beta.14

Major refactoring release that standardizes agent configuration keys to lowercase and introduces the Prometheus agent.

New Features

  • Prometheus Agent: Added new planning agent for strategic task decomposition
  • Agent Display Names: New module for consistent agent name presentation
  • Platform Binary Verification: Enhanced publish workflow with binary verification steps
  • Model Cache Warning: Show warning toast when model cache is unavailable

Breaking Changes (Internal)

  • Agent Keys Normalized to Lowercase: All agent configuration keys are now lowercase throughout the codebase
    • Affected: schema, plugin, model-requirements, agents utils, prometheus-hook
    • Migration added for backward compatibility

Bug Fixes

  • Windows Binary: Fixed segfault by building natively (#1019)
  • Session Preservation: Custom agent now preserved after switching (#1017)
  • Skill Tool: Enforce agent restriction properly (#1018)
  • Look At Tool: Conditionally register when multimodal-looker enabled (#1016)
  • Bash Tool: Use Unix shell syntax on all platforms (#1015)
  • Atlas: Capture stderr from git commands to prevent help text leak
  • Model Resolver: Use first fallback entry when model cache unavailable
  • Doctor: Improved AST-Grep NAPI detection for bunx environments
  • Doctor: Handle file:// protocol for local dev plugin detection
  • LSP: Add data dir to LSP server detection paths (#992)

Documentation

  • Added model configuration section to overview and quick start
  • Renamed Orchestrator-Sisyphus to Atlas
  • Updated multimodal-looker model name and fallback chain

  • c2247ae refactor(agents): add prometheus agent and normalize agent key lookups
  • dfc57d0 refactor(model-requirements): use lowercase agent keys
  • 12c9029 refactor(plugin): use lowercase agent keys throughout
  • 91060c3 refactor(agents): use lowercase config keys in utils
  • 90292db refactor(prometheus-hook): use lowercase config key
  • cc4deed refactor(schema): use lowercase agent config keys
  • 4e42888 refactor(migration): normalize agent keys to lowercase
  • 629a4d3 feat(shared): add agent display names module
  • 8806ed1 feat(publish): add platform binary verification steps
  • e2f8729 @veetase has signed the CLA in #985
  • bee8b37 docs: add model configuration section to overview and quick start to configurations
  • 37e1a06 feat(agents): add aggressive resume instructions to Atlas prompt
  • fc47a7a docs: update multimodal-looker model name and fallback chain
  • 9b12e2a fix(cli): update zai-coding-plan hints to include multimodal-looker
  • 3062277 feat(agents): add zai-coding-plan/glm-4.6v fallback for multimodal-looker
  • 7093583 fix(lsp): add data dir to LSP server detection paths (#992)
  • ec61df8 Merge pull request #913 from carlory/fix-doctor
  • 6312d2d Merge pull request #962 from popododo0720/fix/issues-898-919
  • 810dd93 fix(skill): enforce agent restriction in createSkillTool (#1018)
  • 1a901a5 fix(ci): build Windows binary natively to fix segfault (#1019)
  • f8155e7 fix(session): preserve custom agent after switching (#1017)
  • 39d2d44 fix(tools): conditionally register look_at when multimodal-looker enabled (#1016)
  • 15c4637 fix(hooks): use unix shell syntax for bash tool on all platforms (#1015)
  • 262c711 docs(agents): update AGENTS.md with current commit hash and line counts
  • 599fad0 fix(atlas): capture stderr from git commands to prevent help text leak
  • afbdf69 fix(model-resolver): use first fallback entry when model cache unavailable
  • af9beee @Ssoon-m has signed the CLA in #1014
  • 6973a75 Merge pull request #999 from l3aro/dev
  • c6d6bd1 refactor(models): update agent/category fallback chains
  • 57b1043 fix(agents): use resolved variant from fallback chain instead of requirement default
  • 6dfe091 refactor(atlas): rewrite prompt with lean orchestrator structure
  • 75158ca fix(atlas): register tool.execute.before and pass backgroundManager
  • e16bbbc feat: show warning toast when model cache is not available
  • ab3e622 fix: use cache file for model availability instead of SDK calls
  • f434888 fix: model fallback properly falls through to system default
  • 2c81c8e @l3aro has signed the CLA in #999
  • 3268782 docs: rename Orchestrator-Sisyphus to Atlas
  • be9d6c0 fix(doctor): improve AST-Grep NAPI detection for bunx environments
  • 45fe957 fix(doctor): handle file:// protocol for local dev plugin detection

Don't miss a new oh-my-opencode release

NewReleases is sending notifications on new releases.