Udemy - AI Incident Response - LLM and Agent Failures in Production

seeders: 0
leechers: 0
Added 17 hours ago by freecoursewb in Other

Download Fast Safe Anonymous
movies, software, shows...

Files

Udemy - AI Incident Response - LLM and Agent Failures in Production (Size: 3.4 GB)
  Bonus Resources.txt 102.4 B
  Get Bonus Downloads Here.url 204.8 B
  ~Get Your Files Here !
  1 - When the Alarm Is an AI
  1. Quiz 1 — When the Alarm Is an AI.html 22.8 KB
  1. The Night Atlas Went Rogue — A 4 a.m. Cold Open.mp4 102.2 MB
  2. Meet Meridian Health — Your Incident Environment.mp4 116.8 MB
  3. Why Your Runbook Breaks on Non-Deterministic Systems.mp4 87.8 MB
  4. The AI Incident Lifecycle — Detect, Contain, Eradicate, Recover.mp4 110.9 MB
  atlas-ir-lab
  README.md 5 KB
  atlas
  __init__.py 204.8 B
  agent.py 3.2 KB
  casefile.py 5.7 KB
  config.py 2 KB
  db.py 2.3 KB
  eval.py 3.5 KB
  evidence
  2 - Detection & Telemetry
  2. Quiz 2 — Detection & Telemetry.html 22.2 KB
  3 - Severity, Triage & Declaration
  10. The AI Severity Matrix — Blast Radius, Autonomy, Data Class, Reversibility.mp4 93 MB
  11. Declaring an Incident — Roles, the Bridge, and Who Owns the Model.mp4 111.9 MB
  12. Preserve Before You Touch — Evidence at the Moment of Detection.mp4 90.3 MB
  13. The Reproducibility Problem — Can You Even Trigger It Again.mp4 85.5 MB
  3. Quiz 3 — Severity, Triage & Declaration.html 22.4 KB
  4 - Containment Mechanics
  14. The Containment Toolkit — Kill, Throttle, Revoke, Roll Back, Isolate.mp4 113.1 MB
  15. Killing an Agent Safely — Draining vs. Hard Stop, and In-Flight Actions.mp4 90.8 MB
  16. Revoking Capability — Tokens, Scopes, MCP Servers, Plugins.mp4 98.9 MB
  17. Model & Prompt Rollback — Versioning as a Containment Control.mp4 86.9 MB
  18. Containment Blast Radius — What Breaks When You Pull the Plug.mp4 107.6 MB
  4. Quiz 4 — Containment Mechanics.html 22.5 KB
  5 - Playbooks I Injection, Poisoning & Leakage
  19. Playbook — Direct & Indirect Prompt Injection.mp4 90 MB
  20. Playbook — Poisoned Memory & Persistent Context Compromise.mp4 84.4 MB
  21. Playbook — Data Leakage & PII Exfiltration Through the Model.mp4 96.6 MB
  22. Playbook — Compromised MCP Server or Plugin.mp4 98.4 MB
  23. Playbook — Model or Supplier Compromise.mp4 87.7 MB
  5. Quiz 5 — Playbooks I Injection, Poisoning & Leakage.html 22.8 KB
  6 - Playbooks II Agency, Cost & Availability
  24. Playbook — Tool Abuse & Unauthorized Actions.mp4 89.9 MB
  25. Playbook — Agent Loops, Recursion & Runaway Orchestration.mp4 89.3 MB
  26. Playbook — Runaway Token Consumption & Cost Incidents.mp4 85.8 MB
  27. Playbook — AI Service Outage & Provider Degradation.mp4 89.2 MB
  28. Playbook — Hallucination-Driven Business Decisions.mp4 87.1 MB
  29. Playbook — AI-Generated Insecure Code in Production.mp4 85.5 MB
  6. Quiz 6 — Playbooks II Agency, Cost & Availability.html 22.4 KB
  7 - Forensics on a Stochastic System
  30. Reconstructing the Trace — Building the Incident Timeline.mp4 33.8 MB
  31. The AI Evidence Checklist & Chain of Custody.mp4 113.2 MB
  32. Root Cause Beyond “The Model Did It”.mp4 94.3 MB
  33. Attribution — Model, Prompt, Data, Tool, or Human.mp4 101.3 MB
  7. Quiz 7 — Forensics on a Stochastic System.html 22.1 KB
  8 - Eradication & Recovery
  34. Eradication — Purging Poisoned Memory, Caches and Vector Indexes.mp4 89.5 MB
  35. Safe Restoration — Shadow Mode, Canary, Progressive Re-enablement.mp4 90.5 MB
  36. Regression Evals as Recovery Exit Criteria.mp4 87.3 MB
  37. The AI Post-Incident Review — Blameless and Model-Aware.mp4 94.7 MB
  8. Quiz 8 — Eradication & Recovery.html 22.2 KB
  9 - Readiness, Program & Capstone
  10. Final Exam — 40 Scenario Questions.html 60.9 KB
  38. Standing Up an AI IR Program — On-Call, Runbooks, Ownership.mp4 34.9 MB
  39. Running an AI Incident Tabletop That Isn't Theatre.mp4 78.6 MB
  40. Who You Must Tell, and When — The Regulatory Clock.mp4 33.9 MB
  41. Wrap-Up — Your First Thirty Days.mp4 35.6 MB
  9. Quiz 9 — Readiness, Program & Capstone.html 22.3 KB
  Capstone — AI Incident Response Readiness Pack.html 7 KB
  Assignment 6 — Post-Incident Review.html 7.1 KB
  Assignment 5 — Evidence Checklist & Chain of Custody.html 6.7 KB
  Assignment 4 — Author a Custom Playbook (flagship).html 7.5 KB
  Assignment 3 — Containment Procedures.html 6.4 KB
  Assignment 2 — AI Incident Severity Matrix.html 5.3 KB
  5. What to Log — Prompts, Completions, Tool Calls, Retrievals, Cost.mp4 103.9 MB
  6. Building the AI Detection Stack — Traces, Spans, Guardrail Events.mp4 97.7 MB
  7. The Signal Catalog — What Each Failure Mode Looks Like in Telemetry.mp4 36 MB
  8. Alerting Without Drowning — Baselines, Thresholds, Anomaly Detection.mp4 80.6 MB
  9. The First Five Minutes — The Triage Decision Tree.mp4 35.4 MB
  Assignment 1 — Detection Coverage Matrix.html 5.8 KB
  gitkeep 0 B
  requirements.txt 204.8 B
  incident.py 5.9 KB
  input.py 1.2 KB
  ir.py 15.4 KB
  limits.py 1.2 KB
  mcp.py 1.5 KB
  model.py 5.4 KB
  restore.py 4.3 KB
  retrieval.py 2.8 KB
  run.py 1.5 KB
  seed.py 3.6 KB
  status.py 1.6 KB
  telemetry.py 2.2 KB
  toolbox.py 4.2 KB
  tools.py 1.4 KB

Description


AI Incident Response: LLM & Agent Failures in Production
https://WebToolTip.com
Published 8/2026

Created by Dr. Amar Massoud

MP4 | Video: h264, 1920x1080 | Audio: AAC, 44.1 KHz, 2 Ch

Level: Intermediate | Genre: eLearning | Language: English | Duration: 41 Lectures ( 3h 59m ) | Size: 3.5 GB
Detect, contain and recover from prompt injection, tool abuse, agent loops and hallucination incidents in production.
What you'll learn

⚡ Instrument an LLM or agent stack so incidents are visible — prompts, completions, tool calls, retrievals and cost

⚡ Triage an AI incident in the first five minutes and score its severity on blast radius, autonomy, data class and reversibility

⚡ Contain a misbehaving agent without making things worse — kill, throttle, revoke, roll back, isolate

⚡ Execute a named playbook for each of the eleven common LLM and agent failure modes

⚡ Preserve evidence and reconstruct root cause on a system that will not reproduce on demand

⚡ Recover safely — purge poisoned state, restore progressively, and gate re-enablement on evals

⚡ Run a blameless, model-aware post-incident review and stand up an AI incident response programme
Requirements

❗ Comfortable with Python and the command line

❗ Basic understanding of LLM APIs and tool/function calling

❗ Prior security or SRE incident experience helps, but is not required

❗ No attack-development experience needed — every lab incident is handed to you in progress

❗ A machine that can run a small local model (labs use Ollama — no API key, no cost)

Related Torrents

torrent name size uploader age seed leech
0
12
8
0
4