Update docs and add skills.md files

[dilithjay]

• Updated docs to cover providers, kwargs, return keys, CLI, and entry points that were missing or incorrect.
• Added skills/lexoid-cli and skills/lexoid-python SKILL.md files for the lexoid CLI and the Python API, respectively.

[Copilot]

Pull request overview

This PR expands Lexoid’s documentation to better describe provider setup, CLI usage, and Python API return shapes, and adds SKILL.md guides for the CLI and Python integration.

Changes:

• Add SKILL.md guides for lexoid CLI usage and the Python API.
• Update Sphinx docs to include a new CLI page and refresh API/installation/provider details.
• Adjust docs deployment workflow to only deploy from main.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
skills/lexoid-python/SKILL.md	New Python skill guide with setup, API entry points, and recipes.
skills/lexoid-cli/SKILL.md	New CLI skill guide with commands, flags, and common usage patterns.
docs/installation.rst	Clarifies pip install behavior, provider env vars, and optional dependencies (Playwright/LibreOffice/Ollama).
docs/index.rst	Updates overview text, feature list, providers list, and adds cli to the toctree.
docs/cli.rst	New CLI reference page documenting subcommands and common flags.
docs/benchmark.rst	Fixes numbering and expands benchmark configuration descriptions.
docs/api.rst	Expands API reference (parser types, kwargs, return keys, examples) and adds missing entry points.
.github/workflows/deploy_docs.yml	Ensures the deploy job only runs on refs/heads/main.

Comments suppressed due to low confidence (3)

docs/api.rst:85

• pdf_path is currently added whenever as_pdf=True (even when save_dir isn’t provided, it will point to the temporary converted PDF). The docs say it’s only set when save_dir is specified, which doesn’t match lexoid.api.parse behavior.

   * ``pdf_path``: Path to the intermediate PDF generated when ``as_pdf=True`` and ``save_dir`` is specified.

skills/lexoid-python/SKILL.md:67

• token_cost is described as present whenever api_cost_mapping is supplied, but parse() only sets it if the mapping contains an entry for the selected model. Consider noting that token_cost may be absent (or zero) when the model isn’t in the provided mapping.

    "token_usage": {"input": int, "output": int, "total": int, "llm_page_count": int},
    "token_cost": {...},         # only when api_cost_mapping is supplied
    "parsers_used": [str, ...],  # which parser ran per chunk

skills/lexoid-python/SKILL.md:69

• This says pdf_path is only present when as_pdf=True and save_dir is set, but the implementation adds pdf_path whenever as_pdf=True (it may point to a temp file if save_dir isn’t provided). Please update this return-shape note to match actual behavior.

    "pdf_path": str,             # only when as_pdf=True and save_dir is set
}

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 8 out of 8 changed files in this pull request and generated 5 comments.

[pramitchoudhary]

For example, query:

Sandbox/runtime path: <simple virtual .venv path>

For the given document, it defaults to,

result = parse(
            str(pdf_path),
            parser_type="STATIC_PARSE", or OCR
            framework="pdfplumber",
            retry_on_fail=False,
        )

Should we not consider setting the default config to?

parse("doc.pdf", parser_type="AUTO", router_priority="accuracy", retry_on_fail=True)