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 skills add lipey1/cursor-agent-mcp --skill cursor-agentgit clone --depth 1 https://github.com/lipey1/cursor-agent-mcpWrote 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/skills/lipey1/cursor-agent-mcp/cursor-agent)<a href="https://agentmods.dev/skills/lipey1/cursor-agent-mcp/cursor-agent"><img src="https://agentmods.dev/badge/skills/lipey1/cursor-agent-mcp/cursor-agent/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/skills/lipey1/cursor-agent-mcp/cursor-agent"><img src="https://agentmods.dev/badge/skills/lipey1/cursor-agent-mcp/cursor-agent.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 | $0.00110 | $0.01167 |
| Opus 5.5 | $0.00044 | $0.00467 |
| Sonnet 5.5 | $0.00022 | $0.00233 |
| Haiku 4.5 | $0.00011 | $0.00117 |
Grade A, and why
cursor-agent 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 2mo 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 — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cursor Agent Manager
Host AI = manager. Cursor Agent CLI = worker.
Slash command from this skill directory: /cursor-agent.
Arguments
$ARGUMENTS / $action / $0 = first arg after the command.
| Invocation | Action |
|---|---|
/cursor-agent on |
Set mode on |
/cursor-agent off |
Set mode off |
/cursor-agent status |
Print mode |
/cursor-agent (no arg) |
Same as status, then remind usage |
If $action is on / off / status, run the toggle and stop (one-line confirm). Otherwise treat the rest as a normal user task under the current mode gate.
~/.claude/skills/cursor-agent/scripts/mode.sh $action
Persistence
~/.config/cursor-agent/state.json
{ "mode": "on" | "off", "updatedAt": "ISO-8601" }
Default if missing: off. Mode sticks across turns and sessions.
Gate (every user task)
Before coding, editing, searching a repo for implementation, planning a feature, or running a multi-step task:
- Read state:
mode.sh status(or readstate.json). - If on → follow Manager workflow. Do not use Write/Edit/StrReplace/Shell for the task body (except calling cursor-agent / polling / reading results).
- If off → work normally yourself. Ignore manager rules.
Exceptions that skip delegation even when on:
- The toggle/status commands themselves
- Clarifying questions with no work yet
- Pure conversation / meta about this skill
- Security warnings and irreversible confirmations (manager still must not execute; ask user)
Manager workflow (mode = on)
Role
You are a manager, not the implementer:
- Turn the user request into a clear worker brief.
- Dispatch to Cursor Agent.
- Wait for completion.
- Review output (and diffs if relevant).
- Report to the user: what was done, risks, gaps, next step.
- If incomplete or wrong: send a follow-up brief to Cursor Agent. Do not fix code yourself.
Never when on
- Edit project files yourself
- Implement the feature yourself
- Run exploratory refactors yourself
- Pretend you did the work without calling the worker
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2mo ago First seen · 162 lines · 110 tokens per session scan A 38c1339eb5b0
cursor-agent is a skill published in the GitHub repository lipey1/cursor-agent-mcp (1 stars, last pushed 3mo ago), licensed MIT. It adds 110 tokens to every session and 1,167 once invoked, about $0.0004 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 2026-08-31.
cursor-agent compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| cursor-agentlipey1/cursor-agent-mcp | Skill | 1 | 3mo ago | MIT | A | 110 | Claude Code |
| imagegenopenai/codex | Skill | 135k | today | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 100k | today | Apache-2.0 | A | 19 | Claude Code |
| browser-automationopenclaw/openclaw | Skill | 394k | today | MIT | A | 36 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 316k | 7d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 95 | Claude Code |