docs: generate llms-full.txt for LLM accessibility

[Wenzel]

Fixes #355

Summary

• Adds a single shell step in the docs CI pipeline that concatenates all documentation pages (in SUMMARY.md order) into a single llms-full.txt file
• The file is served at https://intel.github.io/tsffs/llms-full.txt
• Allows LLMs to ingest the full TSFFS documentation in a single request, without crawling ~77 individual HTML pages

Changes

Single addition to .github/workflows/docs.yml after mdbook build:

grep -oP '\([^)]+\.md\)' src/SUMMARY.md | tr -d '()' | while read -r f; do
  echo; cat "src/$f"
done > book/html/llms-full.txt

Why this approach

• No new tools or dependencies — pure shell, runs in the existing CI environment
• Zero ongoing maintenance — file regenerates automatically on every doc change
• ~213 KB / ~6000 lines — fits in a single large LLM context window
• Correct ordering — follows SUMMARY.md chapter order (intro → setup → config → harnessing → fuzzing → tutorials → developer)

Test plan

• Tested generation locally — output verified to be well-formed and complete
• CI passes (docs build + linkcheck)

🤖 Generated with Claude Code