Skip to content

[Phase 1] Agent-native discovery and compatibility #33

Description

@mrgoonie

Depends on

Objective

Make SkillX equally useful to humans and agents for discovery, inspection, and compatibility reasoning.

Scope

  • Compatibility-aware search/detail UI and API.
  • Declared vs verified support.
  • Structured CLI JSON output.
  • Remote read-only MCP.
  • WebMCP read tools as progressive enhancement.
  • Markdown variants of public content routes.
  • Copy Markdown, llms.txt, llms-full.txt, canonical/sitemap work.

Implementation tasks

  • Add --compatible <runtime> search filter.
  • Add skillx inspect.
  • Add skillx check --target.
  • Standardize CLI JSON/error envelope/exit codes.
  • Add compatibility badges with evidence details.
  • Build remote MCP package/server.
  • Implement search_packages, get_package, get_release, check_compatibility, get_kit.
  • Add WebMCP feature detection and read tools.
  • Implement route-level Markdown renderers.
  • Add Link: rel=alternate/canonical behavior where appropriate.
  • Add Copy page as Markdown.
  • Add authorization-aware Copy SKILL.md.
  • Add /llms.txt and /llms-full.txt.
  • Add publisher/harness landing docs.
  • Add parity tests across API/CLI/MCP/Markdown.

Acceptance criteria

An agent can discover a package, inspect requirements, see evidence, and understand incompatibility without scraping HTML.

Browser support for WebMCP is optional; normal web remains fully functional.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions