docs: update comparison for Agent Skills open standard and fair MCP framing

[ReifStanfield]

• Retitle §2 Anthropic Skills → Agent Skills (SKILL.md); reflect multi-host open standard (Cursor, Copilot, Claude Code, Gemini CLI) with links to agentskills.io and anthropics/skills
• Add convergence note in intro clarifying Skillware is not pivoting to SKILL.md
• Soften §1 MCP bullets: acknowledge local stdio option; add complementary- layers sentence (protocol vs in-process capability)
• Update Executive Summary Matrix: column label, Primary Goal cell, and Model Compatibility cell for the Agent Skills column
• Align README Comparison blurb with §2 wording

Description

Type of Change (Matches Issue Templates)

• Skill Proposal: New Skill (Contains manifest.yaml, skill.py, and instructions.md)
• Bug Report Fix: Non-breaking change which fixes an execution error or framework bug
• Doc Fix: Documentation Update
• Framework Feature / RFC Updates: Core Framework Update (Changes to base_skill.py, loader.py, etc.)

Checklist (all PRs)

• My code follows the Agent Code of Conduct.
• I have run python -m flake8 . and pytest tests/ locally (or the subset relevant to this change).
• CHANGELOG.md updated under [Unreleased] if this PR changes user-visible behavior.
• examples/README.md is updated if this PR adds, renames, or removes a runnable script under examples/.

New or updated skill (complete only if this PR adds or changes a skill under skills/)

Skip this section for framework-only, documentation-only, or other PRs that do not touch the skill registry.

Bundle & metadata

• Skill lives at skills/<category>/<skill_name>/ (copied from templates/python_skill/ or equivalent).
• manifest.yaml has name, version, description, valid parameters, and constitution.
• manifest.yaml includes issuer with real name and email (not template placeholders).
• Optional: short_description in manifest (~80 chars) for skillware list.
• Optional: issuer.github and issuer.org set when applicable.
• requirements and env_vars are documented when the skill needs them.

Logic, cognition, and UI

• skill.py is deterministic Python (no arbitrary LLM-generated code paths).
• instructions.md explains when and how to use the skill.
• card.json is present and its issuer matches manifest.yaml (name and email at minimum).

Tests & loader

• test_skill.py covers execution and schema expectations.
• SkillLoader.load_skill("<category>/<skill_name>") succeeds (or missing deps are documented).

Documentation & catalog

• docs/skills/<skill_name>.md exists or is updated (ID, Issuer, usage).
• docs/skills/README.md lists the skill with ID and Issuer.

Constitution & Safety (if adding or modifying a skill)

Related Issues

Fixes #123

[rosspeili]

@ReifStanfield, thanks for #123 / #141, the COMPARISON and README updates look good and match the issue scope.

Before merge, two small things:

1. README.md conflict: Your branch predates recent main README edits. Rebase onto latest ARPAHLS/main and keep upstream README.md everywhere except the Comparison line:

   Skillware differs from the Model Context Protocol (MCP), and Agent Skills (SKILL.md) in several ways:

2. §3 one-liner (optional but appreciated): §3 Antigravity also mentions SKILL.md, easy to blur with §2. Add a single sentence after the Antigravity intro paragraph in COMPARISON.md, e.g.:

   This is distinct from the open Agent Skills (SKILL.md) standard in §2, Antigravity targets IDE procedural memory, not host-agent skill packs.

   One line only, no other §3 changes needed.

Push after rebase and we’ll merge. 🙏

[rosspeili]

@ReifStanfield Thanks again for taking the dive into this. The updates on 40f7e16 look good and read sound to me. The convergence note, §1 MCP wording, §2 Agent Skills framing, matrix cells, README line, and §3 disambiguation all match the issue and our earlier feedback. You defend Skillware clearly without turning it into a format war, which is what we wanted.

Two tiny additions and we can merge:

1. Token Economy, In COMPARISON.md, add one clause to the existing Skillware bullet, contrasting pre-compiled execute() with hosts that load full skill instructions or generate code at runtime. For example: "Hosts using Agent Skills often load full SKILL.md instruction context into the model and may rely on the host to run or generate code, Skillware on the other hand keeps reasoning in instructions.md but runs work in pre-reviewed skill.execute() instead." Wording is up to you, one short sentence is enough.

2. CHANGELOG, Add under [Unreleased] → Documentation, eg. Updated COMPARISON.md and README for Agent Skills (SKILL.md) and fairer MCP framing ([Docs]: Light refresh of COMPARISON.md #123).

Push to the same branch (you don't have to force push, unless you don't want your old commits to be part of the history) and we will merge once CI is green. Nice work overall <3