Literate AI

Repository documentation

Course evidence and limits

These lessons show edited transcript excerpts from real local sessions. They are not unedited screen recordings. Narration is synthetic (Samantha as LitAI, Daniel as Sam); characters are fictional. SSH worker setup is walkthrough-only.

Tested sessions

Framework fixes and video tooling were developed against source base 360efaceecc6b03f78c943f7259ce7317c27be02; the containing Git commit identifies the exact course and tool changes. This is development-release material, not a claim that the new video command already exists in an older published wheel.

The separate build --from-accepted-source continuation on the original 1.0.1 demo still reported source_cache.runtime_absent even with the same model selector. This path is not shown as a successful step; its cause remains under investigation and is not covered by the two confirmed defect repairs above.

Review and reproduction

The media receipt detects changed files and source evidence. It is not a signature or independent proof that narrative claims are correct. Re-execute the documented demo workflows in disposable projects when refreshing for a new release; update this ledger and manifests before rendering. Keep selected model inputs identical between build and verification. Never substitute successful compilation for independent acceptance, exact test counts, packaging, or release evidence.

TinyXML2 is zlib-licensed; this package shows command/output excerpts, not its implementation source. See its license. Character artwork and prompts are retained in assets/; no real person is depicted.

Read the Markdown sourceConference homePublished 1.3.0 · Patch 1.3.1 in progress · Development 1.4