From copilot
Harsh adversarial code review. Assumes the diff is wrong until proven right. Uses a shipped custom agent profile.
How this command is triggered — by the user, by Claude, or both
Slash command
/copilot:adversarial-reviewThis command is limited to the following tools:
The summary Claude sees in its command listing — used to decide when to auto-load this command
# /copilot:adversarial-review
Same input grammar as `/copilot:review`, but uses the adversarial agent profile.
## Behavior
1. Parse `$ARGUMENTS` exactly like `/copilot:review`.
2. **Install the custom agent on first run.** Check whether `~/.copilot/agents/adversarial-review.agent.md` exists. If not, copy it from `${CLAUDE_PLUGIN_ROOT}/agents/adversarial-review.agent.md`:
Never overwrite an existing file — the user may have edited it.
3. Capture diff via `capture-diff.sh` (same as /copilot:review).
4. Resolve model (same default: codex).
5. Invoke `copilot --agent=adversarial-revie...Same input grammar as /copilot:review, but uses the adversarial agent profile.
$ARGUMENTS exactly like /copilot:review.~/.copilot/agents/adversarial-review.agent.md exists. If not, copy it from ${CLAUDE_PLUGIN_ROOT}/agents/adversarial-review.agent.md:
mkdir -p ~/.copilot/agents
if [ ! -f ~/.copilot/agents/adversarial-review.agent.md ]; then
cp "${CLAUDE_PLUGIN_ROOT}/agents/adversarial-review.agent.md" ~/.copilot/agents/
echo "Installed adversarial-review.agent.md to ~/.copilot/agents/"
fi
Never overwrite an existing file — the user may have edited it.capture-diff.sh (same as /copilot:review).copilot --agent=adversarial-review --model "$RESOLVED" -p --silent < diff.The built-in code-review agent is calibrated to be helpful. Sometimes you want a reviewer that defaults to suspicious. That's this command.
After first run, you can edit ~/.copilot/agents/adversarial-review.agent.md to adjust:
The plugin will not overwrite your edits on subsequent runs.
npx claudepluginhub behroozazarkhalili/copilot-plugin-cc --plugin copilot/adversarial-reviewRuns adversarial Codex review challenging implementation approach, design choices, tradeoffs, and assumptions on git-scoped codebase. Supports --wait, --background, --base <ref>, --scope, focus args.
/adversarial-reviewRun a Gemini review that challenges the implementation approach and design choices
/adversarial-reviewRun a Copilot review that challenges the implementation approach and design choices
/adversarial-reviewRuns an adversarial Pi review that challenges implementation approach, design choices, tradeoffs, and assumptions. Returns raw review output from the companion script.
/adversarial-reviewRun a Qwen review that challenges the implementation approach and design choices