Ten courses, grades 6 to 12. Cybersecurity and AI in every one.
Real Windows and Linux machines in the browser, a tutor that makes them show their work, and a transcript line you sign. The AI column counts the lessons where the learner runs, builds, breaks or audits an AI (14 of 34 live today, 252 of 651 planned), not the ones where Dot merely helps: that is all of them. One seat, every course. $720 a year or $60 a month, inside a Texas Education Freedom Account with $1,280 to spare.
Every course below is a full-year plan. Approved counts the hours a Pigeon reviewer has signed off; live is what a learner on 2026-27 can open today. The numbers come from the same records our lesson authors work in, so they move the moment an approval lands.
| Course (Texas name) | Grade | Credit | Planned | Approved · the done metric | Live | AI in this course | Strands · fills | 2026-27 |
|---|---|---|---|---|---|---|---|---|
| Systems & Cyber Basics I What a computer is doing, who it talks to, files and permissions, Python I, and what a model is: the live curriculum's front half with a no-shame computer-basics on-ramp. | 6 | — | 5 h of 140 | 5 h approved (100%) | 14 lessons live today | 11 of 72 lessons: Guess the next word: what an AI model really does · What an AI can see: context, cost and speed · Is the AI actually good? Writing your first eval | SystemsCyberAI Technology applications, grades 6-8 | Available |
| Systems & Cyber Basics II The first alert, boot and services, networking fundamentals, DNS and the web, Python II, and the first tool-using model. | 7 | — | 4 h of 140 | 4 h approved (100%) | 9 lessons live today | 15 of 72 lessons: The anatomy of an alert, and what the AI judge adds · How an AI agent uses a tool (and who really runs it) · Inside an AI agent: the loop and the keys it holds | SystemsCyberAI Technology applications, grades 6-8 | Available |
| Pre-Cyber Performance, base rates, prompt injection, the home network, packet capture, hashing, timelines and phishing: the bridge year and the Summer Cyber Camp source. | 8 | — | 1 h of 140 | 1 h approved (100%) | 3 lessons live today | 8 of 72 lessons: Tricking an AI: instructions hiding in data (prompt injection) · Where sensitive computers live · What a model is for, and isn't | CyberAI Technology applications, grades 6-8 · Summer Cyber Camp | Planned |
| Foundations of Cybersecuritywith AI harnesses, tools & MCP servers Run a model, build a harness, give it tools and MCP servers; malware behaviors, identity, encryption and incident response on real Windows and Linux machines. | 9 | 1.0 | 1 h of 140 | 1 h approved (100%) | 1 lesson live today | 16 of 72 lessons: What did the AI actually do? Auditing an agent from the machine's own evidence · Boot a brain · Tokens, temperature, and why answers wobble | CyberAI Texas cybersecurity program of study | Coming 2027-01 |
| Computer Science Ibuild with AI, review like a pro Specs, tests, building with AI and reviewing its code, plus the CS I core: types, control, arrays, subroutines, style. Texas counts CS I + II as the two LOTE credits. | 10 | 1.0 | 3 h of 140 | 3 h approved (100%) | 6 lessons live today | 56 of 72 lessons: Write a spec a machine can follow · Write the tests first, then let an AI build it · Review the AI's code like a pro | CSAI LOTE credit 1 of 2 · Texas cybersecurity program of study | Planned |
| Networking Routing, segments and firewalls, cloud networking, web application security, threat intelligence, and a network defense plan shown in Observer. | 10 | 1.0 | 0 h of 140 | 0 h approved (0%) | 0 lessons · not yet released | 14 of 72 lessons: The OSI model, labelled on a real capture · The port map: every service, its port, its risk · IPv4 and IPv6, side by side | CyberAI Texas cybersecurity program of study | Planned |
| AI Systems and Governance Texas name: Independent Study in Evolving/Emerging Technologies Agents built, broken and shrunk; the AI session read in Observer; model supply chain; red team and blue team your own agent; train and measure on a GPU; governance in practice. | 11 | 1.0 | 0 h of 140 | 0 h approved (0%) | 0 lessons · not yet released | 72 of 72 lessons: What a harness is · Choosing a model · Context: what the model sees | AI Independent Study in Emerging Technologies | Planned |
| Computer Science IIship software with an AI pair Data structures, object-oriented design, a second language, algorithms and complexity, and a tool with tests, docs and a release. | 11 | 1.0 | 0 h of 140 | 0 h approved (0%) | 0 lessons · not yet released | 43 of 72 lessons: Design an API · Build the service · Validation at the edge | CSAI LOTE credit 2 of 2 | Planned |
| Digital Forensicsincluding AI sessions Windows, Linux and macOS artifacts, imaging and examining a disk, writing a report someone can act on, and detection engineering. | 11 | 0.5 | 0 h of 70 | 0 h approved (0%) | 0 lessons · not yet released | 7 of 37 lessons: What an AI agent leaves behind · Reconstruct from the sensor · Reconstruct from the disk | CyberAI Texas cybersecurity program of study | Planned |
| Cybersecurity Capstone An MCP server for Pigeon built, run under the sensor and defended; a term running an AI security lab on your own machine; certification prep; the senior capstone. | 12 | 1.0 | 1 h of 140 | 1 h approved (100%) | 1 lesson live today | 24 of 72 lessons: Spec it, build it with AI, test it, ship it · Spec and permission list · Build and run under the sensor | CyberAI Texas cybersecurity program of study | Planned |
Course titles on this site are ours; the Texas course name travels on the transcript line. Computer Science I and II satisfy Texas's two-credit languages-other-than-English requirement on the Foundation plan; selective out-of-state colleges may still want a world language. Courses are aligned to the Texas TEKS and designed to transfer; Pigeon Labs Academy is a curriculum vendor, not an accredited school, and the parent or school of record awards the credit.
Guess the next word, the context window, how an agent asks for a tool and who runs it, instructions hiding in data. A family AI policy before grade 9.
Boot a model on a lab machine, write the forty-line chat loop around it, give it a tool, build a first MCP server, then build real software with AI and review its code like a pro.
Red team and blue team their own agent, verify a model before running it, train a small one on a GPU, write the governance file, and build an MCP server for Pigeon that runs under the sensor.
- $720 a year or $360 a semester on the TEFA marketplace
- Sibling seats 10% off
- Every course, the lab machines, Dot, the record
- Four seats or more
- A coach guide for a parent with no CS background; Pigeon carries the insurance
- Book a demo night
- No CS teacher needed: a facilitator runs it
- Accreditation evidence pack; accommodations mode
- Gradebook export