RepoJournal
Agentic Coding Agentic Coding
72 wires and counting

$ follow Agentic Coding

Keep up with Agentic Coding in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-10-02
stories 181

© 2026 RepoJournal Home Showcase Explore How it works Privacy

$ the-wire · showcase

Goose exports sessions to HTML, Codex gets model-listing auth recovery

By RepoJournal · Filed · About Agentic Coding · Composed from the cited sources · methodology

The day's operator-visible work is concentrated in Codex and Goose: a self-contained HTML session export, an opt-in Guardian V2 comparison mode, and a Copilot token fix that stops revoked credentials from poisoning model lists.

feat(cli): add HTML session export with an interactive viewer aaif-goose/goose

by salman1993

goose session export --format html produces a single self-contained file with the session embedded as base64 JSON, rendered client-side with markdown and syntax highlighting. It ships with a scroll-spy table of contents and semantic tool-call rendering for shell, write, todo_write, and computed edit diffs, so transcripts are now shareable as one portable artifact.

fix(provider): refresh revoked Copilot tokens when listing models aaif-goose/goose

by lifeizhou-ap

Model listing used to reuse revoked cached Copilot tokens until their local expiry; it now invalidates the rejected token on 401/403 and retries once with refreshed credentials. Failed recovery surfaces a clear authentication error instead of a stale-token loop. The same recovery logic is shared with completion requests, and a companion PR adds the same behavior at the completion layer.

Add opt-in Decisions comparison for Guardian V2 openai/codex

by copyberry

Codex adds the disabled-by-default guardianv2_decisions_comparison feature, which scores conversations with Decisions alongside the existing Guardian V2 snapshot classification using the same policy and evidence. You opt in with CODEX_GUARDIAN_DECISIONS_API_KEY, and comparison results or failures never change approval outcomes, so it is safe to enable while you evaluate.

Keep fullscreen prompts bounded and scrollable openai/codex

by copyberry

The fullscreen composer is now capped at two-thirds of the screen, with a minimum of eight rows where the screen allows, so long drafts stay browsable while the transcript keeps its space. Mouse-wheel browsing scrolls without moving the caret or disturbing transcript selection, search, or activity focus.

fix(cli): ensure Enter and Spacebar reliably confirm selection list options google-gemini/gemini-cli

by ugorla-dev

Enter and Spacebar now reliably confirm options in selection lists, radio selects, tool confirmation prompts, and AskUserDialog across terminals, including Windows IDE terminals that lack Kitty Keyboard Protocol support; InputPrompt no longer preempts confirmation keystrokes while a prompt or dialog is active. Companion fixes keep the viewport scroll position stable during streaming and long di...

Quick answers

What shipped in Agentic Coding on October 2, 2026?
The day's operator-visible work is concentrated in Codex and Goose: a self-contained HTML session export, an opt-in Guardian V2 comparison mode, and a Copilot token fix that stops revoked credentials from poisoning model lists. In total, 86 commits, 80 pull requests, and 15 releases landed.
Who contributed to Agentic Coding on October 2, 2026?
18 developers shipped this update, including GitHub Actions, copyberry, jif, adamdotdevin, ugorla-dev, luisfelipe-alt, diegogodinezr, and jesussamuel-byte, and 10 more.
What were the notable Agentic Coding updates?
feat(cli): add HTML session export with an interactive viewer, fix(provider): refresh revoked Copilot tokens when listing models, and Add opt-in Decisions comparison for Guardian V2.