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.
git clone --depth 1 https://github.com/Vanyangyang/cursor-bridgeWrote 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/vanyangyang/cursor-bridge/grok_init)<a href="https://agentmods.dev/commands/vanyangyang/cursor-bridge/grok_init"><img src="https://agentmods.dev/badge/commands/vanyangyang/cursor-bridge/grok_init/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/vanyangyang/cursor-bridge/grok_init"><img src="https://agentmods.dev/badge/commands/vanyangyang/cursor-bridge/grok_init.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00017 | $0.00331 |
| Opus 5.5 | $0.00007 | $0.00132 |
| Sonnet 5.5 | $0.00003 | $0.00066 |
| Haiku 4.5 | $0.00002 | $0.00033 |
Grade A, and why
grok_init 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 1mo 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.
What it actually says
Grok Build Supervisor Initialization
Claude Code entry: /grok-build-supervisor:grok_init [loopback-http-url]. Codex entry: $grok-build-supervisor init [loopback-http-url]. The bare /grok_init spelling is a legacy chat alias, not a portable registered host command. An explicit host-expanded invocation retains the same initialization authority.
This command initializes only the persistent local proxy configuration. It must not create or resume a Grok session, open a TUI, send a prompt, or change Grok Executor Mode.
Interpret the entire trimmed value of "$ARGUMENTS" as follows:
- Empty: call
grok_initwithoutproxyUrlso the Supervisor checks configured local candidates and then performs bounded loopback discovery. - One absolute
http://localhost or loopback URL: callgrok_initwith that exactproxyUrlto verify and persist the user's selection. - Anything else: do not call a tool or change state. Reply only with
/grok_initor/grok_init http://127.0.0.1:<port>.
If initialization returns ready, report the verified host and port briefly. If it returns needs_selection, present only the returned candidates and ask the user to rerun /grok_init with one exact URL. If it fails, report the actionable error and next step. Never claim success from an open TCP port alone: the Supervisor must have completed its HTTP CONNECT verification and persisted the setting.
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.
- 1mo ago Changed · +2 lines 4c446c00eed4
- 1mo ago First seen · 16 lines · 17 tokens per session scan A 46b669b69d0f
grok_init is a command published in the GitHub repository Vanyangyang/cursor-bridge (82 stars, last pushed 10d ago), licensed MIT. It adds 17 tokens to every session and 331 once invoked, about $0.0001 per session on Opus 5.5. 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
decision-audit
Post-build decision ledger and pride gate; audit meaningful choices without changing code until revise.
prompt-eval-debug
Debug any prompt with a tiny eval suite (control, edge, boundary), failure diagnosis, and smallest next change, no blind rewrite.
scoped-audit
Scoped, plan-first parallel audit of a large surface using capped subagents, with verification before reporting and an approval gate before any change.
tdd
Canon TDD (specify, encode, fulfill) with one test at a time, approval gates, and kitchen-cleaning refactor when needed.
thermo-nuclear-code-quality-review
Extremely strict maintainability review hunting for code-judo simplifications, giant files, and spaghetti-condition growth.
address-github-pr-comments
Address review comments on the current GitHub PR.