Stop a run that’s gone wrong
Use supported process controls to pause or terminate an agent. Optional policies can escalate when an incident persists.
Process control coverage ↗CAPABILITY MATRIX
Exactly what can be controlled, and where it cannot.
Control depends on the runtime, the OS, the configuration and the plan. This page is the whole matrix, including the empty cells.
The holes are in the table
What we can do, per runtime.
Every vendor in this category says guardrails. This is the whole matrix, including the empty cells.
| Runtime | Observe | Block before it runs | Kill mid-run | Cost in dollars |
|---|---|---|---|---|
| OpenClaw, NVIDIA NemoClaw, Goosefree forever | ✓ | via proxy | ✓ | ✓ |
| Claude Codethe only pre-execution tool gate | ✓ | ✓ | ✓ | ✓ |
| Codex, OpenCode, Aider | ✓ | via proxy | ✓ | ✓ |
| Qwen, Pi, Grok Build, DeepSeek Harnessper-terminal CLI processes, resolved by argv and cwd | ✓ | via proxy | ✓ | ✓ |
| GitHub Copilotbilled in vendor credits, not USD | ✓ | – | ✓ | – |
| Kimi CLIno model id on disk | ✓ | – | ✓ | – |
| Hermes, Deep Agents, n8n, Antigravity, QM, Exo |
✓ | via proxy | – | ✓ |
| NanoClaw, PicoClawno model id on disk | ✓ | – | – | – |
| Cursorone IDE process holds every session | ✓ | – | – | – |
Eight of the 30 have adapters but are not yet audited into this table: Gemini CLI, Lovable, Cline, OpenHands, OpenWorker, Devin, Grok Bot and Replit Agent. They are observed; the other three columns are unverified or unavailable, so they are absent rather than guessed.
Block before it runs
Claude Code gets a real pre-execution gate: a PreToolUse hook parks the call before the tool fires. Everything routed through the ClawMetry proxy gets spend refused with a 429 before tokens are spent.
Kill mid-run
A gateway kill for OpenClaw, and SIGSTOP then SIGKILL against the process tree for the CLI runtimes, with a guard so a recycled pid is never the one we hit. macOS and Linux. Windows has no SIGSTOP, so pause and resume do not exist there.
Audit after
Every event carries a SHA-256 chain over its immutable fields, per node, with a verifier command. Approval decisions land in an append-only log. Works on every runtime.
The three ways to intervene
Approve before a tool runs
Claude Code’s opt-in pre-tool hook holds matching actions for approval. A denied call stays blocked; the agent can try another approach.
How approval gates work ↗Put a limit on routed spend
Use budget alerts for visibility. Route compatible model traffic through the optional proxy to enforce limits before another call.
Explore budget enforcement ↗Pre-tool gating, mid-run control and model-call budget enforcement are different capabilities. Check the matrix for your runtime.