Crabbox is a remote execution control plane that lets a local command run on managed cloud capacity, an SSH host, or another delegated sandbox while syncing code, streaming output, and recording evidence. It is for maintainers, contributors, teams, AI agents, and automation that need repeatable remote testing, builds, browser checks, or platform-specific validation. The catalogue add-ons provide commands and instructions for controlling Crabbox runs.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/openclaw/crabbox/watchgit clone --depth 1 https://github.com/openclaw/crabboxWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/openclaw/crabbox/watch)<a href="https://agentmods.dev/commands/openclaw/crabbox/watch"><img src="https://agentmods.dev/badge/commands/openclaw/crabbox/watch.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01013 |
| Opus 5 | $0.00000 | $0.00507 |
| Sonnet 5 | $0.00000 | $0.00203 |
| Haiku 4.5 | $0.00000 | $0.00101 |
Grade A, and why
watch scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
watch
crabbox watch runs a command on a warm lease, then watches the local
repository and re-runs the command whenever synced files change. Each iteration
goes through the normal crabbox run pipeline against the same held
lease, so sync fingerprints, run history, results, artifacts, and
attach behave exactly as they do for run. It automates the
manual warm-lease loop described in the performance guide.
crabbox watch -- pnpm test:changed
crabbox watch --id swift-crab -- go test ./...
crabbox watch --debounce 500ms --idle-exit 10m -- pnpm check
crabbox watch --provider hetzner --keep -- make test
Requires an SSH lease provider that advertises the crabbox-sync feature (see
crabbox providers). Delegated and archive-sync providers are
rejected with an error; use crabbox run for those.
Lease lifecycle
- With
--id, watch reuses an existing claimed lease and never releases it; exiting the loop leaves the box running, likecrabbox run --id. - Without
--id, watch acquires a fresh lease (allwarmuplease-creation flags apply) and releases it on every exit path: idle exit, Ctrl-C, and failures. Pass--keepto retain the lease instead.
Change detection
The initial run starts immediately. After that, filesystem events are debounced
(default 250ms) and each batch is qualified against the exact sync universe:
Git tracked and non-ignored untracked files, minus the built-in excludes,
sync.exclude config, and .crabboxignore rules, and limited to the
sync.include whitelist when one is configured. Ignored churn such as
node_modules or build output never triggers a run and never counts as
activity. Edits to .gitignore, .crabboxignore, or repo-local config are
picked up live. Newly created directories are watched recursively; symlinked
directories outside the repository root are not followed.
Runs never overlap. A change that lands while a run is active queues exactly one rerun, which starts from the newest tree once the active run finishes. A non-zero remote exit is a normal iteration result and the loop keeps watching; watcher, lease, and sync failures end the loop. One Ctrl-C exits and applies the lease lifecycle rules above.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 88 lines · 0 tokens per session scan A d5c313555e6d
watch is a command published in the GitHub repository openclaw/crabbox (1,368 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,013 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.