Web Standards Archeologist

A specialized protocol for standards engineers and AI agents to trace spec prose back to its historical "Why". It bridges the gap between modern specification fragments and the 20-year history of GitHub issues, Bugzilla reports, and W3C mailing list discussions.

The Problem

Specification source files (like WHATWG's html/source) are massive (130,000+ lines). Standard git blame often only shows the last editorial cleanup or infrastructure update, hiding the actual normative rationale that might date back to 2006.

The "Chain of Evidence" Protocol

This skill defines an archeological protocol for:

1. URL/Fragment -> Source Line: Using heuristics to find where a #fragment is actually defined in the source (even if generated by Bikeshed/Wattsi).
2. Deep Blame: Using git log -L to trace the history of a specific line through refactors and moves.
3. Spec Call Graph: Identifying callers and callees of a specific algorithm across standards (via WebDex) and within a spec.
4. Link Extraction: Identifying and categorizing references to:
   • GitHub: PRs and Issues.
   • W3C Bugzilla: Historical bug reports.
   • SVN Revisions: Direct search links for lists.w3.org archives.
   • Browser Bugs: Chromium, WebKit, and Gecko trackers.

Usage with Gemini CLI

This repository is optimized for use with the Gemini CLI.

1. As an Agent Skill (Recommended)

Add this repository as a specialized skill that Gemini can activate on-demand when it needs to perform spec archeology.

# From within the web-archeologist directory
gemini skills link .

Gemini will now automatically discover the Web Standards Archeologist skill based on the description in SKILL.md and use the activate_skill tool when relevant to your requests.

2. Manual Context Injection

You can inject the protocol into any specific prompt using the @ command:

gemini "Analyze this fragment: https://html.spec.whatwg.org/#main-fetch @SKILL.md"

3. Persistent Workspace Context

To make this protocol available across your entire workspace, import it into your GEMINI.md file:

# My Project Context
...
@/path/to/web-archeologist/SKILL.md

Supported Standards & Engines

• Web Standards: HTML, CSS, DOM, Fetch, URL, etc.
• Chromium: Blink rendering engine (third_party/blink).
• WebKit: WebCore and JavaScriptCore.
• Mozilla: Gecko rendering engine.