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/marcusquinn/aidevops/auto-browsegit clone --depth 1 https://github.com/marcusquinn/aidevopsWhat 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 | $0.00018 | $0.00457 |
| Opus 5 | $0.00009 | $0.00229 |
| Sonnet 5 | $0.00004 | $0.00091 |
| Haiku 4.5 | $0.00002 | $0.00046 |
Grade A, and why
auto-browse 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 yesterday.
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
Run the auto-browse workflow for authorized browser operations and web data mining.
Arguments: $ARGUMENTS
Read .agents/workflows/auto-browse.md, then execute that workflow with the supplied arguments.
Invocation Patterns
| Pattern | Example | Behaviour |
|---|---|---|
| Concrete task | /auto-browse "download invoices from the billing portal" |
Infer scope, ask only materially blocking questions, then plan or run |
| Bare intake | /auto-browse |
Start a guided intake conversation before browsing or writing private state |
| Private workflow review | /auto-browse --review ~/.aidevops/.agent-workspace/auto-browse/<id> |
Inspect an existing private run folder and recommend next action |
| Shareable plan | /auto-browse --plan "create a reusable directory lead exporter" |
Produce a sanitized todo/auto-browse/... plan, not a private account workflow |
Non-negotiables
- Do not browse, log in, create profiles, or touch private browser state until the workflow has target, authorization, allowed actions, and safety policy.
- Store account/site/client-specific workflows, cookies, storage state, downloads, and generated private agents under private aidevops user data, not repo
todo/. - Store only sanitized generalized briefs and implementation plans in
todo/. - Require explicit final confirmation for account creation, posting/messaging, payments/checkouts, destructive/admin actions, anti-detect/proxy/VPN use, and expanded personal/private data extraction.
Related
.agents/workflows/auto-browse.md— canonical workflow.agents/tools/browser/auto-browse.md— tool-routing and subagent design.agents/tools/browser/browser-automation.md— browser tool selection matrix
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.
- yesterday First seen · 45 lines · 18 tokens per session scan A aa72f2b48c37
auto-browse is a command published in the GitHub repository marcusquinn/aidevops (391 stars, last pushed yesterday), licensed MIT. It adds 18 tokens to every session and 457 once invoked, about $0.0001 per session on Opus 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
build-agent
Build an AI agent on Cloudflare using the Agents SDK.
adopt-roadmap.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
landing-page-generator.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
perf-profile.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
github-epic-menu.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
landing-page-generator-executor.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.