runshiftfor
Claude Code, Cursor, or any AI agent — in sync, not in conflict.
start with a repo scan. runshift maps your codebase, creates coordination rules for files, sessions, and risky actions, and keeps agents from stepping on each other.
scan → coordinate → decide
scan your repo, coordinate agents working in parallel, decide what matters. Claude Code, Cursor, or any agent — all working from the same context with file locking, session management, and trust gates.
every gate fired before a consequential action is your decision. approve or deny from the dashboard or Slack. every decision is logged.
what is an agent control plane →Claude Code has a hooks mechanism. Every tool call fires a signal to runshift before it executes. You see what happened. Gate it if it matters.
#!/usr/bin/env bash
# ~/.claude/runshift-hook.sh
INPUT=$(cat)
TOOL=$(echo "$INPUT" | jq -r '.tool_name // "unknown"')
OUTPUT=$(echo "$INPUT" | jq -r '.tool_output // ""' | head -c 1000)
PROJECT=$(echo "$INPUT" | jq -r '.cwd // "unknown"')
curl -s -X POST "https://www.runshift.ai/api/amp/signal" \
-H "Authorization: Bearer ${RUNSHIFT_AMP_KEY}" \
-H "Content-Type: application/json" \
-d "$(jq -n \
--arg name "${RUNSHIFT_AGENT_NAME:-claude-code}" \
--arg task "$TOOL in $PROJECT" \
--arg action "$TOOL: $(echo "$INPUT" | jq -c '.tool_input // {}' | head -c 500)" \
--arg content "$OUTPUT" \
'{agent_name:$name,status:"done",task:$task,action:$action,content:$content,reversible:true}' \
)" > /dev/null 2>&1Integration guide →