# reevesagents (npm · reevesagents)

Spawn, read and steer AI coding CLIs (Claude Code, Codex, Kimi and more) in local tmux windows.

- Trust score: 81/100 (high trust)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-25

## Components

- npm · `reevesagents`: 81/100 (this document), [markdown](https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents.md), [page](https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents)

## Channel facts

- Registry: `npm`
- Package: `reevesagents`
- Version: `2.0.0`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-10-25.

- **Supply Chain Security**: 86/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - Install-script check failed: the install command fetches or executes arbitrary code (inline_eval). An install hook runs on every machine, in CI, and on transitive installs, whether or not you ever run the server.
  - 43 of 140 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 100/100
  - Source repository is publicly reachable at the declared URL.
  - Cryptographically verified build provenance (signed, bound to mertkayacs/reevesagents).
  - Clear OSI-approved license (Apache-2.0).
  - Actively maintained (last published 6 days ago).
  - Publishes a security disclosure policy (SECURITY.md).
- **Schema Quality & AI Usability**: 82/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Tool/resource definitions use about 2746 tokens (~83/item across 33 items; 31 tools + 2 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 37/100
  - Stability observed for 11 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 83/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 50% of tool parameters carry a description.
- **Tool Safety**: 75/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - 0 of 7 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "kill" implies "kill" and declares no destructiveHint at all, which the MCP spec reads as destructive by default.
  - An AI judge read all 33 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### How do I install the reevesagents MCP server?

reevesagents runs locally as an npm package, launched with npx -y reevesagents. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add mertkayacs-reevesagents -- npx -y reevesagents
```

### Cursor

```json
{
  "mcpServers": {
    "mertkayacs-reevesagents": {
      "command": "npx",
      "args": [
        "-y",
        "reevesagents"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "mertkayacs-reevesagents": {
      "command": "npx",
      "args": [
        "-y",
        "reevesagents"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add mertkayacs-reevesagents -- npx -y reevesagents
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "mertkayacs-reevesagents": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "reevesagents"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add mertkayacs-reevesagents --command npx --arg -y --arg reevesagents
```

### Hermes

```yaml
mcp_servers:
  mertkayacs-reevesagents:
    command: "npx"
    args: ["-y", "reevesagents"]
```

### Netclaw

```json
{
  "McpServers": {
    "mertkayacs-reevesagents": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "reevesagents"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add mertkayacs-reevesagents -t stdio -c npx -a -y reevesagents
```

### Other

```json
{
  "mcpServers": {
    "mertkayacs-reevesagents": {
      "command": "npx",
      "args": [
        "-y",
        "reevesagents"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-10-25 (score 81, +1)

No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-23 (score 80, +1)

No change was recorded against any check on this day. Stability & Change Management went from 27 to 30. That category is still filling its 30-day observation window: 8 days of observed history at the previous scan, 9 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-20 (score 79, +1)

No change was recorded against any check on this day. Stability & Change Management went from 17 to 20. That category is still filling its 30-day observation window: 5 days of observed history at the previous scan, 6 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-19 (score 78, 0)

- [security regression] Stability: 0.13 → unverified
- [security regression] Tool safety: pass → unverified
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional] Package version: 1.9.1 → 2.0.0

### 2026-10-18 (score 78, +16)

- [security improvement] Malware scan: unverified → pass

### 2026-10-17 (score 62, −15)

- [security regression] Malware scan: pass → unverified
- [functional] Package version: 1.9.0 → 1.9.1

### 2026-10-16 (score 77, +1)

- [security regression] Tool safety: pass → unverified
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional improvement] Stability: unverified → 0.07
- [functional] Package version: 1.8.2 → 1.9.0

### 2026-10-14 (score 76, +30)

- [security improvement] Injection markers: unverified → pass
- [security] First check of Judged manipulation: pass
- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] Schema quality: unverified → 100
- [functional improvement] MCP protocol: unverified → pass
- [functional] First check of Schema quality: fail
- [functional] First check of Tool coverage: 49
- [functional] First check of Destructive annotations: 0
- [functional] First check of Schema quality: pass
- [functional] First check of Schema quality: good

## MCP tools (31)

### `spawn` (~305 tokens)

Start a CLI agent in its own tmux window and return its agent_id (use it with send_text, send_key, read, kill) and run_id. Omit run_id to add the agent to this MCP session run, created on the first spawn so a chain of spawns stays in one run; pass run_id to add it to an existing run. Call list_providers for valid provider keys.

Input parameters:

- `auth_mode` (string): How the CLI authenticates. Blank uses the provider default.
- `effort` (string): Reasoning effort, for providers that support it. Blank uses the provider default.
- `extra_args` (array): Extra flags appended verbatim to the launch, for provider options ReevesAgents does not model. Each flag and its value is a separate item, e.g. ["--remote-control"].
- `model` (string): Model id. Blank keeps the provider default.
- `name` (string): Run name when starting a new run.
- `nickname` (string)
- `permissions` (string): ask prompts for each action; skip runs autonomously. Blank uses the global default.
- `provider` (string, required): Provider key, e.g. cc, codex, kimi, qwen, opencode, hermes.
- `run_id` (string): Add to this run. Omit to start a new run.
- `task` (string): Initial prompt pasted into the agent on startup.
- `working_dir` (string)

### `kill` (~46 tokens)

Stop one agent and close its tmux window. Pass an agent_id from spawn or list. Once it is ended you can remove its record with delete.

Input parameters:

- `agent_id` (string, required)

### `stop` (~51 tokens)

Stop a whole run: end every agent in it and tear down its tmux session. Pass a run_id from spawn or list. Once ended, remove it with delete_run.

Input parameters:

- `run_id` (string, required)

### `send_text` (~57 tokens)

Paste text into one agent at its prompt. This does NOT submit it: follow with send_key enter to send. Use it to give a running agent a new instruction.

Input parameters:

- `agent_id` (string, required)
- `text` (string, required)

### `send_key` (~64 tokens)

Press one key in an agent: enter, escape, tab, space, up, down, left, right, backspace, or ctrl-c. Send enter after send_text to submit the pasted text.

Input parameters:

- `agent_id` (string, required)
- `key` (string, required)

### `interrupt` (~28 tokens)

Press ctrl-c in one agent to interrupt whatever it is doing.

Input parameters:

- `agent_id` (string, required)

### `read` (~66 tokens)

Read one agent's recent terminal output, ANSI-stripped and secret-redacted. Use it to see what the agent printed or is waiting on. Defaults to the last 20 lines.

Input parameters:

- `agent_id` (string, required)
- `lines` (number): How many lines, default 20.

### `list` (~53 tokens)

List every live run and the agents in each, with their ids, status, provider, and tmux info. Start here to find a run_id or agent_id for the other tools. Ended runs are in list_history.

### `list_providers` (~43 tokens)

List the CLI providers this machine can launch: id, display name, whether it is installed, aliases, and known models. Pass an id as the provider for spawn.

### `list_tools` (~78 tokens)

List the helper tool CLIs on this machine (Perplexity, Google Colab, Hugging Face, Ollama): what each is for, whether it is installed, and its state (signed in, or reachable for Ollama), plus an example command and a docs link. They are not spawn providers; run them from your own shell.

### `request_approval` (~82 tokens)

Ask for approval before a risky action. Creates a pending approval that a human or another agent clears with resolve_approval, and returns it with its approval_id.

Input parameters:

- `action` (string, required)
- `agent_id` (string, required): The agent asking for approval.
- `details` (object)
- `risk` (string)
- `summary` (string, required)

### `resolve_approval` (~55 tokens)

Approve or deny one pending approval by its approval_id (from list_approvals). An optional note records why.

Input parameters:

- `approval_id` (string, required)
- `decision` (string, required)
- `note` (string)

### `check_approval` (~38 tokens)

Read one approval's current status (pending, approved, denied, or expired) by its approval_id.

Input parameters:

- `approval_id` (string, required)

### `list_approvals` (~55 tokens)

List approval requests, optionally filtered by run_id or status (pending, approved, denied, expired). Use it to find an approval_id to resolve.

Input parameters:

- `run_id` (string)
- `status` (string)

### `list_history` (~30 tokens)

List archived run history: runs that have ended or gone stale. Live runs are in list, not here.

### `delete` (~36 tokens)

Permanently remove one ended agent's record. Kill the agent first; a still-running agent is rejected.

Input parameters:

- `agent_id` (string, required)

### `delete_run` (~41 tokens)

Permanently remove one ended run, archiving it to history. Stop the run first; a still-running run is rejected.

Input parameters:

- `run_id` (string, required)

### `delete_history` (~31 tokens)

Permanently remove one archived run history record by its id (from list_history).

Input parameters:

- `id` (string, required)

### `open` (~107 tokens)

Open a run or agent in tmux for the user: switch the attached tmux client to a run (its reeves tab) or to a specific agent window. Works for any run, whether this agent started it or not, so you can surface any run for the user on request. No effect when nobody is attached.

Input parameters:

- `id` (string, required): A run id or an agent id (from list). A run id opens its reeves tab; an agent id opens that agent window.

### `doctor` (~58 tokens)

Check this machine for running agents: tmux availability, which provider CLIs are installed, and the state directories. Returns each check with an ok/warn/fail status and a detail string. Run it when a spawn fails to see what is missing.

### `reap` (~120 tokens)

Sweep for zombie agents and orphan tmux sessions: ends any agent whose tmux window has died or that is older than max_lifetime_ms when that is set above 0 (see get_config), and kills reeves-*/reevesweb_* sessions no live record owns. Returns each reaped agent with the reason (window-gone or lifetime-exceeded) and its age, plus the killed orphan session names. This runs on its own in the background; call it to force an immediate sweep, e.g. after a crash or before relying on list.

### `get_config` (~76 tokens)

Read the global reevesagents settings: peek_interval_ms, peek_lines, max_depth, max_agents (the per-run agent cap), ready_delay_ms, max_lifetime_ms (auto-reap agents older than this; 0 disables), default_permissions, and language. Check max_agents here when a spawn is rejected for hitting the cap.

### `set_config` (~256 tokens)

Change one or more global settings and save them. Pass only the fields to change. Counts must be positive integers (ready_delay_ms and max_lifetime_ms may be 0); default_permissions is ask or skip; language is a supported code such as en or tr. Example: raise the per-run agent cap with { "max_agents": 20 }, or auto-kill agents running longer than an hour with { "max_lifetime_ms": 3600000 }. Returns the saved settings.

Input parameters:

- `default_permissions` (string): Permission mode used when a spawn omits one.
- `language` (string): UI language code for the TUI and web UI, e.g. en, tr.
- `max_agents` (number): Max agents per run.
- `max_depth` (number): Spawn recursion cap.
- `max_lifetime_ms` (number): Auto-reap any agent older than this many ms on the next sweep. 0 disables lifetime reaping.
- `peek_interval_ms` (number): Milliseconds between output peek polls.
- `peek_lines` (number): Lines captured per peek.
- `ready_delay_ms` (number): Delay before the startup prompt is injected.
- `web_theme` (string): Web UI appearance.

### `list_presets` (~43 tokens)

List saved run presets: reusable agent-team templates, each with a root agent and workers. Launch a whole team at once by passing a preset name to start_preset.

### `save_preset` (~115 tokens)

Save a live run's agent line-up as a reusable preset under a name, so the same team can be relaunched later with start_preset. Pass a run_id from list. Re-saving the same name updates it. Note: a captured agent's auth_mode and effort are not stored (they default).

Input parameters:

- `description` (string)
- `name` (string, required): Preset name; characters that are not filename-safe are replaced.
- `run_id` (string, required): The run whose agents become the preset (from list).

### `start_preset` (~92 tokens)

Launch a new run from a saved preset (see list_presets), starting its root agent and every worker at once. Optionally override the run name and working_dir. Returns the new run and its agents.

Input parameters:

- `name` (string, required): Preset name to launch (from list_presets).
- `run_name` (string): Run name; defaults to the preset name.
- `working_dir` (string)

### `delete_preset` (~38 tokens)

Permanently delete a saved preset by its name (from list_presets). Does not affect any running run.

Input parameters:

- `name` (string, required)

### `list_hosts` (~51 tokens)

List the AI CLIs on this machine that can host the reevesagents MCP and whether reevesagents is attached to each. Use it to see where this control surface is wired before attach_host or detach_host.

### `attach_host` (~77 tokens)

Attach the reevesagents MCP to one host CLI by its key (from list_hosts), or to every installed host when key is omitted. After attaching, that CLI can drive agents through these same tools.

Input parameters:

- `key` (string): Host key from list_hosts, e.g. cc or codex. Omit to attach all installed hosts.

### `detach_host` (~49 tokens)

Detach the reevesagents MCP from one host CLI by its key (from list_hosts). The CLI keeps running but can no longer drive agents through reevesagents.

Input parameters:

- `key` (string, required)

### `install_skills` (~114 tokens)

Install the reevesagents Agent Skill so skill-aware CLIs (Claude Code, Codex, Kimi, OpenCode) learn to drive reevesagents on their own. Writes one SKILL.md to the two shared skill directories under the home dir. Pass no arguments; call with status omitted to install, or read the returned rows to see where it landed. The user must restart their CLIs to pick it up.

Input parameters:

- `status` (boolean): When true, only report where the skill is installed without writing anything.

## Diagnostics

Captured diagnostic sections: Provenance, Install scripts, Dependencies. The full working is on the page: https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents#diagnostics

## Score history

- 2026-10-25: 81
- 2026-10-24: 80
- 2026-10-23: 80
- 2026-10-22: 79
- 2026-10-21: 79
- 2026-10-20: 79
- 2026-10-19: 78
- 2026-10-18: 78
- 2026-10-17: 62
- 2026-10-16: 77
- 2026-10-15: 76
- 2026-10-14: 76
- 2026-10-13: 46
- 2026-10-12: 46

## Common questions

### What is the reevesagents MCP server?

reevesagents is an MCP server listed in the public MCP registry as io.github.mertkayacs/reevesagents. Spawn, read and steer AI coding CLIs (Claude Code, Codex, Kimi and more) in local tmux windows. This page covers its npm package (reevesagents).

### Is the reevesagents MCP server safe to use?

reevesagents scores 81 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 25 October 2026. It declares an install script we flagged as high risk. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the reevesagents MCP server expose?

reevesagents exposes 31 tools: spawn, kill, stop, send_text, send_key, and 26 more. Their descriptions and schemas cost roughly 2,355 tokens of context every time the server is loaded.

### Is the reevesagents MCP server still maintained?

reevesagents is still listed as active in the MCP registry. We last reached this channel on 25 October 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the reevesagents MCP server under?

reevesagents declares the Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- npm package: https://www.npmjs.com/package/reevesagents
- Socket report: https://socket.dev/npm/package/reevesagents
- Repository: https://github.com/mertkayacs/reevesagents
- Website: https://reevesagents.mertkayacs.com/
- Changelog RSS feed: https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents.xml
- Changelog JSON feed: https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents.json
- HTML version of this page: https://verifymcp.io/servers/mertkayacs-reevesagents/reevesagents