docs

name	docs
description	Update EggPdf documentation — README, docs/ wiki pages, site/, llms.txt, CLAUDE.md — to reflect current implementation
user-invocable	true
allowed-tools	Read, Edit, Write, Bash, Grep, Glob

Docs Updater

Audit and update all EggPdf documentation to stay in sync with the implementation.

Scope

File/Dir	Purpose
README.md	GitHub repo front page + NuGet description
docs/*.md	GitHub Wiki pages (14 files)
site/	GitHub Pages static site
site/llms.txt	AI/LLM discovery (site-scoped)
llms.txt	AI/LLM discovery (repo root)
CLAUDE.md	Claude Code development rules
spaces/README.md	HuggingFace Space metadata

Steps

1. Audit current state

Check for common staleness signals:

# Version mismatches
grep -r "version\|Version" src/EggPdf/EggPdf.csproj src/EggPdf.Service/Program.cs site/index.html README.md | grep -E "\"[0-9]+\.[0-9]+\.[0-9]+\""

# Wrong Docker image name (should be eggspot/eggpdf)
grep -r "eggpdf/service\|eggpdf/cli" docs/ README.md site/

# Outdated benchmark placeholder
grep -n "Phase 1\|will be added" README.md

# HuggingFace short_description length (must be ≤ 60 chars)
python3 -c "
import re
content = open('spaces/README.md').read()
m = re.search(r'short_description: (.+)', content)
if m: print(len(m.group(1)), 'chars:', m.group(1))
"

2. Fix version numbers

• site/index.html: Update softwareVersion in JSON-LD and v{x.y.z} in hero badge
• src/EggPdf.Service/Program.cs: Use typeof(EggPdf.HtmlToPdf).Assembly.GetName().Version instead of hardcoded string

3. Fix Docker image names

Correct name is eggspot/eggpdf:latest (check .github/workflows/docker.yml to confirm). Replace any eggpdf/service:latest or eggpdf/cli references in docs/ and README.md.

4. Update benchmarks

If <!-- BENCHMARK_START --> in README.md has a placeholder, replace with current numbers from CLAUDE.md:

| Simple page (h1 + p) | **13 µs** | 20 KB |
| Invoice (table + styles) | **86 µs** | 85 KB |
| Large table (100 rows) | **1.2 ms** | 939 KB |

5. Check llms.txt files

Both llms.txt (root) and site/llms.txt should have:

• Current install command
• Working code examples
• Performance numbers
• Link to full AI guide: https://raw.githubusercontent.com/eggspot/EggPdf/main/llms.txt

6. Check spaces/README.md

• short_description must be ≤ 60 characters
• app_port must be 8080
• API example must use the correct HF Space URL

7. Check docs/ wiki pages

For each changed feature, verify the corresponding wiki page is accurate:

• Endpoint names, request/response shapes, option names
• Code examples actually compile (no removed APIs)
• Version requirements (e.g., "requires EggPdf 1.0+")

8. CLAUDE.md

Verify:

• Project structure section matches actual src/ and tests/ layout
• Test count comment is approximately correct (dotnet test --list-tests | wc -l)
• Benchmark table reflects current numbers
• Key files list is accurate

What NOT to change

• Don't document stub/unimplemented features as if they work (e.g. image rendering, full signing)
• Don't add fictional API surface that doesn't exist in code
• Don't change docs/*.md content without verifying against actual implementation in src/

After making changes

Run a quick sanity check:

# Build still passes
dotnet build src/EggPdf/EggPdf.csproj -c Release --no-restore

# No broken links in README (basic check)
grep -o 'https://[^)\"]*' README.md | sort -u