Add commands

Author: SimonSchubert

assets/commands/doc-engine.md

@@ -0,0 +1,107 @@
+# TAGLINE
+
+Zero-config Markdown to PDF documentation engine using Typst
+
+# TLDR
+
+**Generate a PDF** from README.md with zero configuration
+
+```doc-engine build```
+
+**Convert a specific file** and override title and author
+
+```doc-engine build CONTRIBUTING.md -o guide.pdf -t "Guide" -a "Team"```
+
+**Use a different template** and accent color
+
+```doc-engine build --template modern --accent teal```
+
+**Check for Markdown errors** without generating PDF
+
+```doc-engine build --dry-run```
+
+**Install via pipx**
+
+```pipx install doc-engine-cli```
+
+# SYNOPSIS
+
+**doc-engine** build [INPUT_FILE] [options]
+
+**doc-engine** info
+
+**doc-engine** --help | --version
+
+# DESCRIPTION
+
+**doc-engine** is a CLI tool that converts Markdown files into professionally styled PDF documents using Typst as the rendering backend. It auto-detects README.md, extracts metadata from Git, and produces high-quality technical or academic reports complete with cover page, table of contents, and syntax-highlighted code.
+
+It requires no LaTeX, Pandoc, or manual configuration. The tool supports five built-in templates and custom accent colors. It can also perform source error checking with precise line/column reporting.
+
+# PARAMETERS
+
+**build** [_INPUT_FILE_]
+
+> Convert the Markdown file to PDF. Defaults to auto-detecting README.md in the current directory.
+
+**-o**, **--output** _file.pdf_
+
+> Output file path. Default: <input>_doc.pdf
+
+**-t**, **--title** _title_
+
+> Document title. Defaults to first # heading or Git-derived.
+
+**-a**, **--author** _name_
+
+> Author name. Defaults to `git config user.name`.
+
+**--template** _name_
+
+> Template: academic (default), modern, minimal, technical, book.
+
+**--accent** _color_
+
+> Accent color (hex or named: blue, teal, rose, etc.).
+
+**--bib** _file.bib_
+
+> Bibliography file for IEEE-style references section.
+
+**--dry-run**
+
+> Validate Markdown and report errors without producing PDF.
+
+**--open**
+
+> Open the generated PDF after creation.
+
+**--no-branding**
+
+> Omit "doc-engine" attribution from footer.
+
+# EXAMPLES
+
+Basic zero-config usage in a project directory:
+
+```doc-engine build```
+
+Generate with custom options:
+
+```doc-engine build docs/api.md -o api.pdf --template technical --accent "#7c3aed"```
+
+# CAVEATS
+
+Images in Markdown are rendered only as alt text (remote images not downloaded or embedded). Complex layouts may require reviewing the output PDF. Requires Python 3.10+ and the Typst Python bindings (installed automatically).
+
+# SEE ALSO
+
+[typst](/man/typst)(1), [pandoc](/man/pandoc)(1), [weasyprint](/man/weasyprint)(1)
+
+# RESOURCES
+
+```[Source code](https://github.com/leonardosalasd/doc-engine-cli)```
+
+```[Documentation](https://github.com/leonardosalasd/doc-engine-cli#readme)```
+
+<!-- verified: 2026-07-07 -->

assets/commands/index.txt

@@ -1723,6 +1723,7 @@ dnsx.md
 do-release-upgrade.md
 do.md
 doas.md
+doc-engine.md
 docbook2html.md
 dockdiver.md
 docker-build.md
@@ -3923,6 +3924,7 @@ kyanos.md
 kyma.md
 l123.md
 l2ping.md
+l9gpu.md
 labcoat.md
 lambo-new.md
 lambo.md

assets/commands/l9gpu.md

@@ -0,0 +1,77 @@
+# TAGLINE
+
+GPU observability agent with Kubernetes and Slurm workload attribution
+
+# TLDR
+
+**Install** the Python package
+
+```pip install l9gpu```
+
+**Monitor NVIDIA GPUs** and emit OTLP telemetry
+
+```l9gpu nvml_monitor --sink otel --cluster my-cluster```
+
+**Quick test** without sending data
+
+```l9gpu nvml_monitor --sink stdout --once```
+
+**Monitor AMD GPUs**
+
+```l9gpu amd_monitor --sink otel --cluster my-cluster```
+
+**Monitor Intel Gaudi**
+
+```l9gpu gaudi_monitor --sink otel --cluster my-cluster```
+
+**Run via systemd** or Kubernetes DaemonSet (see Helm)
+
+# SYNOPSIS
+
+**l9gpu** _subcommand_ [options]
+
+Common subcommands: nvml_monitor, amd_monitor, gaudi_monitor, slurm_monitor and others.
+
+# DESCRIPTION
+
+**l9gpu** is the core CLI for Last9 GPU Telemetry. It runs as a per-node agent that collects hardware metrics (NVML, amdsmi, hl-smi) from GPUs and enriches them with workload context from Kubernetes or Slurm before exporting standard OpenTelemetry (OTLP) metrics and logs.
+
+This enables per-pod, per-job, per-user GPU accounting without building custom pipelines. It supports NVIDIA, AMD, and Intel Gaudi GPUs and integrates with any OTLP-compatible backend.
+
+The package also provides related health check tooling.
+
+# PARAMETERS
+
+**--sink** _otel|stdout|file_
+
+> Destination for telemetry (otel recommended for production).
+
+**--cluster** _name_
+
+> Cluster identifier attached to emitted data.
+
+**--once**
+
+> Run a single collection cycle and exit (useful for testing).
+
+Other options and subcommands are listed via `l9gpu --help`.
+
+Configuration can also be supplied via TOML files.
+
+# CAVEATS
+
+Requires appropriate vendor libraries or tools (NVML for NVIDIA, etc.) and Python 3.10+. Primarily intended for cluster operators; not a replacement for DCGM or full Prometheus exporters. Emits OTLP; a collector is typically used downstream for storage/visualization.
+
+# SEE ALSO
+
+[otel](/man/otel)(1), [dcgm](https://github.com/NVIDIA/DCGM)
+
+# RESOURCES
+
+```[Source code](https://github.com/last9/gpu-telemetry)```
+
+```[Homepage](https://last9.io/gpu-observability/)```
+
+```[Documentation](https://last9.io/docs/)```
+
+<!-- verified: 2026-07-07 -->