Skip to content

Remove legacy tool-wrapper fallback after projection helper is required #547

@chubes4

Description

@chubes4

Problem

PR #538 migrated selected DMC workspace and GitHub model tools to Data Machine's ability-native projection helper, but intentionally kept legacy wrapper registration as a fallback when the helper is unavailable and left bespoke alias/composed wrappers in place.

That is the right migration shape, but the fallback should be tracked so DMC does not permanently maintain two model-tool registration paths.

Desired outcome

Once DMC requires a Data Machine version that includes datamachine_register_ability_tool(), remove the legacy fallback registration for tools that can be projected directly from abilities. Keep bespoke wrappers only where they add real model-specific behavior.

Acceptance criteria

  • Raise or document the minimum Data Machine dependency that guarantees the projection helper exists.
  • Remove fallback BaseTool registration for tools fully covered by ability-native projections.
  • Keep model-facing tool names stable unless a deliberate migration note is added.
  • Document which remaining wrappers are intentionally bespoke and why.
  • Regression tests prove projected DMC tools are offered to the model and execute through canonical abilities.

Related

AI assistance

  • AI assistance: Yes
  • Tool(s): OpenCode (openai/gpt-5.5)
  • Used for: Boundary audit follow-up and tracker drafting; Chris remains responsible for prioritization.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions