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/benchgit clone --depth 1 https://github.com/openclaw/crabboxWhat 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 | $0.00000 | $0.01204 |
| Opus 5 | $0.00000 | $0.00602 |
| Sonnet 5 | $0.00000 | $0.00241 |
| Haiku 4.5 | $0.00000 | $0.00120 |
Grade A, and why
bench 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 — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
bench
crabbox bench records and reports local benchmark timing observations. It is a
local evidence workflow, not a provider leaderboard.
Benchmark rows are append-only JSONL records. They wrap the existing
TimingReport payload with local comparison context such as command
fingerprint, command display text, provider family/kind, repo fingerprint, and
cold/warm state when known. Crabbox only writes this ledger when explicitly
requested.
Default store:
<CrabboxStateDir()>/timings.jsonl
CrabboxStateDir() uses $XDG_STATE_HOME/crabbox when set; otherwise it uses
the user config directory's crabbox/state directory.
Record from a run
Use run --timing-record to append the final timing report from a real run:
crabbox run --timing-record=default -- pnpm test
crabbox run --timing-record ./bench/timings.jsonl --provider aws -- pnpm test
Plain crabbox run remains non-recording by default.
Run and record a benchmark
bench run runs the same command for each selected provider and repeat, then
records observations in the benchmark store:
crabbox bench run --providers aws,hetzner --repeats 3 -- pnpm test
crabbox bench run --provider aws --cold -- go test ./...
The command uses the same execution path as crabbox run, so provider setup,
sync, command streaming, timing, and normal cleanup behavior stay centralized.
If any provider or repeat fails, Crabbox continues the remaining attempts and
exits non-zero after printing the number of recorded observations. Use
--store <path> to write a specific JSONL store; the default is the local state
store.
Record an existing timing JSON payload
bench record ingests one saved TimingReport JSON object from a file or
stdin:
crabbox bench record --timing-json timing.json --command "pnpm test" --cold
--command or command args after -- set the command display text and command
fingerprint used for grouping:
crabbox bench record --timing-json timing.json -- pnpm test
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 · 152 lines · 0 tokens per session scan A 60e94d300758
bench is a command published in the GitHub repository openclaw/crabbox (1,343 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,204 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 2026-08-30.
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.