Prompt

After longer turns, suggests 3 next tasks for the project with a Do it button each, and a Dismiss.

Plan-mode turns run on Sonnet; once you approve the plan, the build starts as a fresh turn on your session's model (Opus).

Hides SecChain's secret values in prompts, tool results and instructions before the model reads them, through secchain mask.

Concise prose and YAGNI code, installed into the project AGENTS.md. /endsession writes Rules, Next and Todo, commits, pushes, then stops.

Unofficial: lets Claude drive macOS apps in the background through the computer-use engine bundled with the ChatGPT desktop app, without taking over your cursor.

Delete a file and its ghost haunts your prompt, whispering fragments of its lost code until you restore it. Seances, exorcisms, poltergeists and secrets.

入力欄の上の帯で、作業に合わせて Clawd（二次創作版）が踊り、使用量も表示する

/q holds prompts until the running turn ends and sends them one per turn; a band above the prompt lists, edits, reorders and pauses them

Share a Claude Code session with a link. Teammates join from their own Claude Code and talk to it: one session does the work, everyone sees every turn live, and the host approves what runs on their machine. Sharing goes through this project's public server unless you set your own (README: Privacy and the public server).

Shows images pasted into the prompt as thumbnails above it

A line above the prompt for each pull request Claude opens: its GitHub workflows with a progress bar, then ready to merge, the failing job, or what blocks it.

Automated multi-agent code review cycle. Spawns reviewers in parallel (plus Codex when available), applies fixes per CLAUDE.md policy, admits a commit only when a reviewer saw exactly what it records, and asks before a push, pull request, merge or release you did not request, unless you let it go further.

A live side pane that maps a long Claude Code session: goal, current path, possible detours, decisions, open questions, checkpoints, working files and what to resume next. It observes on its own and asks before changing your intent.

Counts the time you really spend with Claude Code (typing and Claude working, not an idle terminal) and shows it in the status line and /claude-time.

Re-explain the last answer as ASD-STE100 text, a diagram, an HTML page or an explainer video, and score answers for STE style.

A quote button on every chat message, an edit button on your own, and a 5 second undo for prompts sent while a turn runs.

Saves a 3 line card when you stop (doing, next step, open question) and shows it above the prompt next time

Blocks AI sounding text in social media writing (X replies, X posts, LinkedIn posts) and injects the writing rules.

Stops fix loops: when the same error returns after 3 fix attempts, tells you and Claude to stop and rethink

Uncommitted files and time since last commit in the status line, with a nudge after editing turns

Writes a short recap of what you did when you step away, and shows it above the prompt when you come back

Git Diff timeline above the prompt: click a commit to see it, click two to compare them, with their commit messages; a Branches tab compares two branches; a GitHub-style files view shows each file's diff. Works with any git remote.

Status line showing session time, prompts, and tool calls.

Verifies signatures on the instruction files Claude Code loads, meaning CLAUDE.md, skills, and agent definitions, before they reach model context. It re-checks a skill bundle before the Skill tool runs it. Sigstore keyless, offline verification against a pinned trust root, no telemetry.