AI › Module 2 › Lesson 4
Lab — LLM Basics
Pack models APIs, context and tools, and trust boundaries cards — LLM basics file from $AI_LAB only.
Visual · t32_llm_basics_lab
Lab: LLM basics pack. $AI_LAB only. Original Cyberlium.
Opening
LLM basics pack merges API to context to boundaries — paperwork before injection module.
Lessons 2-1–2-3 named models and APIs, context and tools, and trust boundaries. This lab merges three cards into one $AI_LAB LLM basics artifact with M1 ethics cross-reference. No stranger app access — LLM literacy paperwork only. Next: Quiz — LLM Basics Literacy.
1. Lab contract: LLM basics pack
Create $HOME/cyberlium-lab/t32-m02-l04-llm-basics-lab.txt merging models APIs, context and tools, and trust boundaries sections with $AI_LAB toy name and M1 ethics cross-reference.
Optional: one-line note which toy chatbot you used — fictional names OK.
Command guide
Try these commands — Lab contract: LLM basics pack
═══ TOOLS & WEBSITES ═══ Browse / read these (authorized learning only — stay in YOUR lab / program scope)
OpenAI docs — https://platform.openai.com/docs/ Hugging Face security — https://huggingface.co/docs/hub/security MITRE ATLAS — https://atlas.mitre.org/
═══ INSTALL ═══
Linux (Debian/Ubuntu):
Command — copy this
sudo apt install python3
macOS:
Command — copy this
brew install python3
Windows: Download https://python.org/downloads/
═══ LINUX / macOS ═══
Command — copy this
export AI_LAB=${AI_LAB:-$HOME/cyberlium-lab/t32-ai}
python3 - <<'PY'
import os
lab = os.environ.get('AI_LAB', os.path.expanduser('~/cyberlium-lab/t32-ai'))
path = os.path.join(lab, 'notes', 'llm-basics-lab-summary.md')
open(path, 'w').write('
'.join([
'# LLM Basics Lab Summary',
'## Completed (YOUR lab)',
' - context-and-tools-literacy.md reviewed',
' - trust-boundaries-checklist.md drafted',
'## Key takeaways',
' - Models complete text; tools expand blast radius',
' - Untrusted context is an injection surface',
'## Refusals',
' - No stranger API scraping; no jailbreak weaponization',
]))
print(f'Wrote {path}')
PYCommand — copy this
grep -E 'Refusals|trust-boundaries' "$AI_LAB/notes/llm-basics-lab-summary.md"
═══ WINDOWS ═══
Command — copy this
Get-Content $HOME/cyberlium-lab/t32-ai/notes/trust-boundaries-checklist.md | Select-String untrusted
Primary tools to practice this lesson: grep, python3. Reference sites: OpenAI docs (https://platform.openai.com/docs/); Hugging Face security (https://huggingface.co/docs/hub/security); MITRE ATLAS (https://atlas.mitre.org/). Run every command in the box — install first, then the usage lines — only on YOUR lab / program scope.
2. Cross-check refusal boundaries
Grep pack for API key secrets, stranger endpoint URLs, jailbreak references, malware prompts — redact or delete. Pack stays documentary literacy.
Update pack if you rename lab toy — note new bot name.
3. Lock the proof
chmod 600 on the pack. Quiz next — then Direct Injection.
LLM basics literacy feeds prompt injection lessons in M3.
4. What you ship: LLM basics pack for $AI_LAB
Merged API, context, boundaries sections. $AI_LAB named. NO stranger endpoints. chmod 600.
5. What you record before the next lesson
Date. LLM basics pack path. $AI_LAB named. File t32-m02-l04-llm-basics-lab.txt chmod 600.
6. Wrong vs right: stranger SaaS vs YOUR toy LLM
Worked failure — same MSF word, opposite target. Right never needs a café Wi-Fi or classmate laptop.
Wrong
Include live API key in pack. Skip trust boundaries section.
Right
Write YOUR LLM basics pack for $AI_LAB. chmod 600. Next: Quiz — LLM Basics Literacy.
Mission: freeze YOUR LLM basics pack
1) Merge M2 literacy sections. 2) Confirm zero stranger refs or key secrets. 3) Link M1 ethics cross-ref. 4) chmod 600.
Stuck? Ask Cyberlium AI Mentor
One governed LLM pack beats ten unauthorized API screenshots.
Knowledge Check
APPLY: This lab requires:
Multiple choice
Knowledge Check
APPLY: True or False: LLM basics pack should reference $AI_LAB toy chatbot.
True or False
Knowledge Check
APPLY: LLM basics pack should include:
Multiple choice