Skip to content
Back to search

call-e-integrations

call-e integrations for phone calling and real-world outreach across OpenClaw, Codex, and other agent platforms.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/CALLE-AI/call-e-integrations with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/CALLE-AI/call-e-integrations'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/calle-ai--call-e-integrations/badge.svg)](/listings/calle-ai--call-e-integrations)

HTML

<a href="/listings/calle-ai--call-e-integrations"><img src="/api/listings/calle-ai--call-e-integrations/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

**The fastest path — paste this into any AI agent (Claude Code, Codex, Cursor, and more):**

Install CALL-E for me: https://open.heycall-e.com/document/mcp-archive/CALL-E-installation-guide.md

Your agent handles the rest.

**SDK:** Follow the [TypeScript or Python quickstart](#sdk) with an explicit recipient.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 29 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "name": "call-e-claude",
  3. "owner": {
  4. "name": "CALLE AI"
  5. },
  6. "metadata": {
  7. "description": "CALL-E plugins for Claude Code."
  8. },
  9. "plugins": [
  10. {
  11. "name": "calle",
  12. "source": "./packages/claude-plugin/plugin",
  13. "description": "Use CALL-E phone call workflows from Claude Code through the calle CLI.",
  14. "version": "0.2.4",
  15. "author": {
  16. "name": "CALLE AI",
  17. "url": "https://github.com/CALLE-AI"
  18. },
  19. "category": "Productivity",
  20. "tags": [
  21. "call-e",
  22. "calle",
  23. "mcp",
  24. "phone-call",
  25. "voice-call"
  26. ]
  27. }
  28. ]
  29. }
README.md · README · 585 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <div align="center">
  2. # CALL-E Integrations
  3. **CALL-E is your AI agent for getting phone work done.**
  4. Tell CALL-E your goal, and it handles the phone task end-to-end: it plans, calls, adapts in real time, follows through, and improves along the way.
  5. Use CALL-E directly, or integrate it into agents, platforms, and business systems through Skills, Plugins, SDKs, or APIs.
  6. New users receive 100 free credits (US$1) when they sign up. See the [website FAQ](https://www.heycall-e.com/) for current pricing and your [dashboard billing details](https://dashboard.heycall-e.com/account/billing) for actual usage charges.
  7. [Website](https://www.heycall-e.com/) · [Docs](https://docs.heycall-e.com/) · [Try on ClawHub](https://clawhub.ai/call-e-dev/phone-call-calle) · [Discord](https://discord.gg/6AbXUzUV8w)
  8. ![npm](https://img.shields.io/npm/v/@call-e/cli?label=%40call-e%2Fcli)
  9. ![Codex](https://img.shields.io/badge/Codex-CALL--E-black)
  10. ![Claude Code](https://img.shields.io/badge/Claude%20Code-CALL--E-orange)
  11. ![Cursor](https://img.shields.io/badge/Cursor-CALL--E-blue)
  12. ![OpenClaw](https://img.shields.io/badge/OpenClaw-ClawHub-purple)
  13. ![Hermes Agent](https://img.shields.io/badge/Hermes%20Agent-ClawHub%20Prompt-green)
  14. ![MCP](https://img.shields.io/badge/MCP-Streamable%20HTTP-blue)
  15. </div>
  16. ## Quick Start
  17. **The fastest path — paste this into any AI agent (Claude Code, Codex, Cursor, and more):**
  18. ```text
  19. Install CALL-E for me: https://raw.githubusercontent.com/CALLE-AI/call-e-integrations/main/docs/install/CALL-E-installation-guide.md
  20. ```
  21. Your agent handles the rest.
  22. **SDK:** Follow the [TypeScript or Python quickstart](#sdk) with an explicit recipient.
  23. ## Contents
  24. - [What is CALL-E?](#what-is-call-e)
  25. - [Capabilities](#capabilities)
  26. - [Get Started](#get-started)
  27. - [Agent Install](#agent-install)
  28. - [MCP](#mcp)
  29. - [SDK](#sdk)
  30. - [API](#api)
  31. - [Supported Regions and Languages](#supported-regions-and-languages)
  32. - [Examples](#examples)
  33. - [Troubleshooting](#troubleshooting)
  34. - [In Development](#in-development)
  35. - [Repository Structure](#repository-structure)
  36. - [Telemetry](#telemetry)
  37. - [Development](#development)
  38. - [Community](#community)
  39. ---
  40. ## What is CALL-E?
  41. CALL-E automates goal-driven phone tasks that scripted voice bots cannot handle.
  42. Traditional calling platforms use prebuilt bots optimized for high-volume, repetitive scripts. CALL-E is different: you describe a goal, and CALL-E figures out how to achieve it over the phone. It handles natural conversation, adapts to unexpected responses, and returns a structured result when the call ends.
  43. This makes CALL-E practical for tasks where a rigid script would fail — appointment confirmations, research calls, follow-ups, lead qualification.
  44. **Call lifecycle:**
  45. ```mermaid
  46. flowchart LR
  47. A["Goal + phone number"] --> B["Plan"]
  48. B --> C{"Details\ncomplete?"}
  49. C -- "Missing info" --> D["Clarify"]
  50. D --> B
  51. C -- "Confirmed" --> E["Dial"]
  52. E --> F["Live conversation"]
  53. F --> G["Structured result\n+ transcript\n+ summary"]
  54. ```
  55. ---
  56. ## Capabilities
  57. | Capability | Description |
  58. | --- | --- |
  59. | **Live Task Progress** | Track a call from planning to completion: status, activity history, outcomes, and next steps |
  60. | **Smart Goal Clarification** | CALL-E asks for missing details — recipient, timing, language, success criteria — before dialing |
  61. | **Managed Call Execution** | Handles number setup, outbound dialing, monitoring, and result capture |
  62. | **Structured Results** | Returns summaries, transcripts, and schema-validated structured data you can act on directly |
  63. | **Scheduled and Batch Calling** | Schedule individual calls or send a batch task to multiple recipients |
  64. | **In-Task Optimization** | Adapts call strategy based on prior attempts within the same task |
  65. | **Real-World Voice Handling** | Manages live pickup, voicemail, call screening, hold, transfers, silence, and interruptions |
  66. | **IVR Navigation** | Detects and navigates IVR menus during outbound calls, using DTMF keypad input when needed to reach the requested department, queue, automated service, or person |
  67. | **Multiple Integration Paths** | Agent plugins, MCP, SDKs, APIs, and enterprise systems |
  68. | **Safety and Governance** | Number governance, rate limits, concurrency controls, blocklists, kill switches, redacted logs, and audit trails |
  69. ## In Development
  70. **Goal-Driven Long Tasks** — CALL-E plans a multi-step task end-to-end: it designs the calling approach, executes the calls, learns from real outcomes, and continuously improves its strategy over time. This goes beyond single calls — CALL-E learns how to achieve each phone-based goal more reliably across attempts. This feature is under active development and not yet generally available.
  71. ## Get Started
  72. Choose the integration path that fits your use case:
  73. | Use case | Integration | Start here |
  74. | --- | --- | --- |
  75. | Use CALL-E inside Claude Code, Codex, Cursor, OpenClaw, Hermes, or any skills.sh agent | Agent install | [Agent Install](#agent-install) |
  76. | Connect any Streamable HTTP MCP client | MCP | [MCP](#mcp) |
  77. | Call CALL-E from a TypeScript or Python SDK | SDK | [SDK](#sdk) |
  78. | Call CALL-E from any backend | Developer API | [API](#api) |
  79. ---
  80. ### Agent Install
  81. **Paste this single prompt into your agent for automatic setup:**
  82. ```text
  83. Install CALL-E for me: https://raw.githubusercontent.com/CALLE-AI/call-e-integrations/main/docs/install/CALL-E-installation-guide.md
  84. ```
  85. The linked guide installs the standalone `calle` 1.0.6 Skill and selects the
  86. host-supported connection path. The initial live acceptance baseline is Codex
  87. MCP; API and other clients identify unexecuted live checks separately.
  88. For manual setup, expand the table below or see the [full install guide](https://github.com/CALLE-AI/call-e-integrations/blob/main/docs/install/install-guide.md).
  89. For the standalone workflow that prefers MCP with API Key fallback under its
  90. [interface selection rules](docs/install/calle.md#choose-the-interface),
  91. see [calle installation](docs/install/calle.md). Business operations and updates
  92. use non-CLI host facilities. MCP setup prefers native host management, with a
  93. limited exception for the current host's official MCP management commands.
  94. The standalone Skill is now `calle` version `1.0.6`. Users retaining the legacy
  95. CLI workflow can install `calle-cli` through the
  96. [compatibility guide](docs/install/skills-sh-skill.md#existing-calle-users).
  97. Plugin-owned `calle` Skills have independent installation and release lifecycles.
  98. ### MCP
  99. CALL-E supports Streamable HTTP MCP with OAuth. The standalone `calle` workflow
  100. uses `https://seleven-mcp-sg.airudder.com/mcp/openagent_oauth` and the current tools
  101. `prepare_call`, `start_call`, `open_call`, `get_call`, and `get_call_transcript`.
  102. Server clarification works through chat or optional cards. Once the latest
  103. preparation is ready and the user has authorized execution, the Agent uses
  104. `start_call` without redundant confirmation. Planning and preview requests do
  105. not authorize dialing. Cancellation requires supported card controls. See the
  106. [calle guide](docs/install/calle.md#choose-the-interface) for routing and recovery.
  107. The legacy CLI and plugin integrations retain the separate Plan/Run contract
  108. below; the standalone `calle` does not use those tools as a fallback.
  109. **Legacy endpoint:**
  110. ```text
  111. https://seleven-mcp-sg.airudder.com/mcp/openagent_oauth
  112. ```
  113. **Transport:** Streamable HTTP
  114. **Legacy tool flow:**
  115. ```mermaid
  116. sequenceDiagram
  117. participant Client as MCP Client
  118. participant CE as CALL-E
  119. participant Phone
  120. Client->>CE: plan_call(goal, phone)
  121. CE-->>Client: plan_id, confirm_token, ready_to_run=true
  122. Note over Client: User confirms intent
  123. Client->>CE: run_call(plan_id, confirm_token)
  124. CE-->>Client: run_id
  125. CE-)Phone: Outbound call
  126. Note over Client,CE: Follow next_step, default first wait ~60s
  127. loop Until terminal status, stop instruction, or confirmation request
  128. Client->>CE: get_call_run(run_id)
  129. CE-->>Client: status, activity, transcript, next_step
  130. end
  131. ```
  132. **Legacy tools:**
  133. | Tool | What it does |
  134. | --- | --- |
  135. | `plan_call` | Creates or refines a call plan. Does **not** place a call. Returns `plan_id`, `confirm_token`, and `ready_to_run`. |
  136. | `run_call` | Starts the planned call. Requires the exact `plan_id` and `confirm_token` from the preceding `plan_call`. **Can place a real phone call.** |
  137. | `get_call_run` | Reads run status, activity, summary, and transcript. Read-only. Follow `next_step`; without guidance, wait ~60 seconds before the first poll, then every 5–10 seconds until terminal. |
  138. <!-- sync-with: docs/mcp/openagent-oauth.md#reliable-terminal-state-workflow -->
  139. Server-directed delays, stop instructions, and retry-confirmation questions
  140. take precedence over the default cadence. Ask the user before a requested
  141. retry; do not place another call automatically. The ~60-second delay is a
  142. polling recommendation, not a completion deadline.
  143. Persist the returned `run_id` and resume `get_call_run` after a local timeout
  144. or restart; do not call `run_call` again. MCP `run_call` does not accept a
  145. `webhook_url`, so MCP clients should poll `get_call_run` for completion.
  146. For OAuth details, tool contracts, setup, and the completion workflow, see
  147. the [MCP guide](https://github.com/CALLE-AI/call-e-integrations/blob/main/docs/mcp/openagent-oauth.md#reliable-terminal-state-workflow).
  148. ### SDK
  149. CALL-E server SDKs are available for TypeScript and Python. Use them in trusted backend services, workers, and automation systems.
  150. **Install:** Python requires version 3.11 or later.
  151. ```bash
  152. # TypeScript
  153. pnpm add @call-e/calle@1.0.1
  154. # Python
  155. pip install calle-ai==1.0.1
  156. ```
  157. **Set your API key, recipient, and workflow key:**
  158. ```bash
  159. export CALLE_API_KEY="<YOUR_CALLE_API_KEY>"
  160. export CALLE_BASE_URL="https://api.heycall-e.com"
  161. export CALLE_EXAMPLE_PHONE="<AUTHORIZED_E164_PHONE>"
  162. export CALLE_IDEMPOTENCY_KEY="<UNIQUE_WORKFLOW_KEY>"
  163. ```
  164. Replace the non-working key placeholder with your key from the [CALL-E dashboard](https://dashboard.heycall-e.com/account/api-keys).
  165. These examples place a real call. Use a number you own or are authorized to call;
  166. for integration testing, follow the [official test-hotline instructions](https://discord.com/channels/1493880186826133504/1495622983253889054/1546414916515401788).
  167. Choose and save a unique workflow key before the first request. Reuse it only for
  168. the same request; a new key can create another call.
  169. The examples below use SDK **1.0.1** and the single-target Calls V2 API.
  170. Existing integrations must [migrate inputs and result handling](https://docs.heycall-e.com/migration)
  171. before upgrading. Historical call-task IDs still use the legacy interface.
  172. | Field | Meaning |
  173. | --- | --- |
  174. | `phone` | Required, one E.164 phone number. |
  175. | `region` | Optional destination country/region code, inferred from the phone. |
  176. | `locale` | Optional spoken-language hint, inferred from the task and supported regional languages. |
  177. Calls require a result schema with a closed object of scalar fields
  178. (`additionalProperties: false`); arrays and nested objects are unsupported.
  179. The idempotency key is required. Missing or conflicting task inputs return
  180. `422 input_incomplete`; collect the requested facts before resubmitting.
  181. **TypeScript:**
  182. ```ts
  183. import { CalleClient } from "@call-e/calle";
  184. const client = new CalleClient({
  185. apiKey: process.env.CALLE_API_KEY!,
  186. baseUrl: process.env.CALLE_BASE_URL!,
  187. });
  188. const created = await client.calls.create({
  189. task: "Say hello in English, allow a brief reply, then politely end the call.",
  190. phone: process.env.CALLE_EXAMPLE_PHONE!,
  191. resultSchema: {
  192. type: "object",
  193. additionalProperties: false,
  194. required: ["summary"],
  195. properties: {
  196. summary: { type: "string" },
  197. },
  198. },
  199. }, { idempotencyKey: process.env.CALLE_IDEMPOTENCY_KEY! });
  200. console.log("Save this Call ID before waiting:", created.id);
  201. const call = await client.calls.waitForResult(created.id);
  202. console.log(call.status, call.callOutcome, call.resultStatus);
  203. console.log(call.result, call.error);
  204. console.log(call.transcript);
  205. ```
  206. **Python:**
  207. ```python
  208. import os
  209. from calle import CalleClient
  210. client = CalleClient(
  211. api_key=os.environ["CALLE_API_KEY"],
  212. base_url=os.environ["CALLE_BASE_URL"],
  213. )
  214. created = client.calls.create(
  215. task="Say hello in English, allow a brief reply, then politely end the call.",
  216. phone=os.environ["CALLE_EXAMPLE_PHONE"],
  217. idempotency_key=os.environ["CALLE_IDEMPOTENCY_KEY"],
  218. result_schema={
  219. "type": "object",
  220. "additionalProperties": False,
  221. "required": ["summary"],
  222. "properties": {
  223. "summary": {"type": "string"},
  224. },
  225. },
  226. )
  227. print("Save this Call ID before waiting:", created["id"])
  228. call = client.calls.wait_for_result(created["id"])
  229. print(call["status"], call["call_outcome"], call["result_status"])
  230. print(call["result"], call["error"])
  231. print(call["transcript"])
  232. ```
  233. **Python exceptions:**
  234. Import the exported exception classes directly from `calle`:
  235. ```python
  236. from calle import (
  237. CalleAPIError,
  238. CalleAuthenticationError,
  239. CalleConnectionError,
  240. CalleRateLimitError,
  241. CalleTimeoutError,
  242. CalleWebhookSignatureError,
  243. )
  244. ```
  245. | Exception | When to handle it |
  246. | --- | --- |
  247. | `CalleAPIError` | API error responses, including HTTP 422. Inspect `status_code` and `code`. |
  248. | `CalleAuthenticationError` | HTTP 401/403; a subclass of `CalleAPIError`. |
  249. | `CalleRateLimitError` | HTTP 429; a subclass of `CalleAPIError`. |
  250. | `CalleConnectionError` | Transport failures; separate from `CalleAPIError`. |
  251. | `CalleTimeoutError` | Request or polling timeout; separate from `CalleAPIError`. |
  252. | `CalleWebhookSignatureError` | Legacy signed-webhook verification only; current webhooks are unsigned. |
  253. Catch authentication/rate-limit subclasses before `CalleAPIError` when handling
  254. them separately. A polling timeout does not cancel an accepted call. For a
  255. complete example that saves the original request and Call ID and resumes polling,
  256. see the [Calls example and recovery guide](https://docs.heycall-e.com/quickstart#run-a-complete-example).
  257. Wait helpers stop when `result_status` is no longer `pending` (`resultStatus`
  258. in TypeScript). `completed` alone does not establish result readiness or business
  259. success. Use the separate `id` for API requests and `call_id` (`callId` in TypeScript)
  260. for Billing lookup.
  261. #### Community SDKs
  262. - [Kotlin / Android SDK](https://github.com/Baklolman69/calle-android-sdk) — Community-maintained Kotlin SDK for integrating CALL-E voice agents into Android and Wear OS apps.
  263. ### API
  264. The CALL-E Developer API provides direct HTTP access for any trusted backend, worker, or workflow system.
  265. **Set credentials:** Replace `<YOUR_CALLE_API_KEY>` with the complete key from the [CALL-E dashboard](https://dashboard.heycall-e.com/account/api-keys).
  266. ```bash
  267. export CALLE_API_KEY="<YOUR_CALLE_API_KEY>"
  268. export CALLE_BASE_URL="https://api.heycall-e.com"
  269. export CALLE_IDEMPOTENCY_KEY="<UNIQUE_WORKFLOW_KEY>"
  270. ```
  271. **Endpoints:**
  272. | Method | Path | Description |
  273. | --- | --- | --- |
  274. | `POST` | `/v2/calls` | Prepare and accept one immediate phone call (HTTP 202). |
  275. | `GET` | `/v2/calls/{call_id}` | Read execution status, result readiness, result, error, and transcript. |
  276. | `GET` | `/v2/calls/{call_id}/events` | List developer-facing call events. |
  277. | `POST` | `/v2/calls/{call_id}/cancel` | Cancel before provider submission; active calls cannot be hung up through this operation. |
  278. The webhook URL is yours. Pass it as `webhook_url` when creating a call. See the [webhooks guide](https://docs.heycall-e.com/webhooks).
  279. **Create a call:**
  280. ```bash
  281. curl --fail-with-body --silent --show-error "$CALLE_BASE_URL/v2/calls" \
  282. --request POST \
  283. --header "Authorization: Bearer $CALLE_API_KEY" \
  284. --header "Content-Type: application/json" \
  285. --header "Idempotency-Key: $CALLE_IDEMPOTENCY_KEY" \
  286. --data '{
  287. "task": "Say hello in English, allow a brief reply, then politely end the call.",
  288. "phone": "<AUTHORIZED_E164_PHONE>",
  289. "result_schema": {
  290. "type": "object",
  291. "additionalProperties": false,
  292. "required": ["summary"],
  293. "properties": {
  294. "summary": {
  295. "type": "string"
  296. }
  297. }
  298. },
  299. "metadata": {
  300. "workflow_run_id": "hello-001"
  301. }
  302. }'
  303. ```
  304. **Read a result:**
  305. ```bash
  306. export CALLE_CALL_ID="<ID_FROM_CREATE>"
  307. curl --fail-with-body --silent --show-error "$CALLE_BASE_URL/v2/calls/$CALLE_CALL_ID" \
  308. --header "Authorization: Bearer $CALLE_API_KEY"
  309. ```
  310. Poll until `result_status` is not `pending`. `available` means `result` is ready;
  311. `unavailable` means stop waiting even when both `result` and `error` are null.
  312. `not_applicable` means inspect cancellation or the technical error.
  313. After a timeout, query the saved ID or replay the unchanged original request/key;
  314. a new key can place another call. See the [retry rules](https://docs.heycall-e.com/calls#idempotency).
  315. **Ready result excerpt:**
  316. <details>
  317. <summary>Example response</summary>
  318. ```json
  319. {
  320. "status": "completed",
  321. "call_outcome": "completed",
  322. "result_status": "available",
  323. "result": { "summary": "The recipient replied hello." },
  324. "error": null,
  325. "transcript": [
  326. { "offset_seconds": 0, "speaker": "bot", "text": "Hello!" },
  327. { "offset_seconds": 4, "speaker": "user", "text": "Hello." }
  328. ]
  329. }
  330. ```
  331. </details>
  332. For authentication, webhooks, and the full reference, see the [API docs](https://docs.heycall-e.com/api-reference).
  333. ---
  334. ## Supported Regions and Languages
  335. Use these country codes with the SDK and API recipient settings.
  336. | Country | Country Code | Calling Code | Languages | Line Region |
  337. | --- | --- | --- | --- | --- |
  338. | United States of America | `US` | +1 | English, Indonesian | Local |
  339. | Canada | `CA` | +1 | English | Local |
  340. | Australia | `AU` | +61 | English | International |
  341. | Bangladesh | `BD` | +880 | English, Bengali | International |
  342. | Brazil | `BR` | +55 | English, Portuguese | International |
  343. | Germany | `DE` | +49 | English, German | International |
  344. | Spain | `ES` | +34 | English, Spanish | International |
  345. | Finland | `FI` | +358 | English, Finnish | International |
  346. | United Kingdom of Great Britain and Northern Ireland | `GB` | +44 | English | International |
  347. | Indonesia | `ID` | +62 | English | International |
  348. | India | `IN` | +91 | English, Hindi, Tamil | International |
  349. | Japan | `JP` | +81 | English, Japanese | International |
  350. | Mexico | `MX` | +52 | English, Spanish | International |
  351. | Malaysia | `MY` | +60 | English, Malay, Mandarin Chinese | International |
  352. | Netherlands | `NL` | +31 | English | International |
  353. | Philippines | `PH` | +63 | English | International |
  354. | Pakistan | `PK` | +92 | English, Urdu | International |
  355. | Poland | `PL` | +48 | English, Polish | International |
  356. | Singapore | `SG` | +65 | English | International |
  357. | Thailand | `TH` | +66 | English, Thai | International |
  358. | Turkey | `TR` | +90 | Turkish | International |
  359. | Viet Nam | `VN` | +84 | English, Vietnamese | International |
  360. **Notes**
  361. - **Local** means calls use a local number in the destination country or calling-code region.
  362. - **International** means calls use an international number. These numbers are intended primarily for development and call testing, and calls through the shared international number pool may cost more.
  363. - **Region not listed?** Use a SIP integration to call that destination. For risk-control reasons, we enable additional International regions for development and testing only after careful review.
  364. ---
  365. ## Examples
  366. Runnable demos are in [examples/](https://github.com/CALLE-AI/call-e-integrations/blob/main/examples):
  367. | Example | What it shows |
  368. | --- | --- |
  369. | [Standard MCP OAuth clients](https://github.com/CALLE-AI/call-e-integrations/blob/main/examples/mcp-oauth-client) | TypeScript and Python clients connecting to CALL-E via standard MCP OAuth over Streamable HTTP. Good starting point for any new MCP client. |
  370. | [CALL-E broker login MCP clients](https://github.com/CALLE-AI/call-e-integrations/blob/main/examples/mcp-broker-client) | TypeScript and Python clients using CALL-E brokered login, local token caching, and MCP HTTP calls. Useful when the environment cannot complete a browser OAuth flow. |
  371. | [Python batch runner](https://github.com/CALLE-AI/call-e-integrations/blob/main/examples/python-batch-runner) | Python JSONL batch runner using `calle` CLI auth state, FastMCP, Rich output, and MCP tool-call metadata. Demonstrates processing multiple call tasks from a file. |
  372. These are starting-point demos, not the canonical SDK or API contract.
  373. ---
  374. ## Troubleshooting
  375. If installation, authentication, or MCP tool verification fails, see the [CALL-E troubleshooting guide](https://github.com/CALLE-AI/call-e-integrations/blob/main/docs/install/troubleshooting.md).
  376. Common issues covered:
  377. - **Cursor network errors** — if setup works in your terminal but fails in the Cursor agent shell, see the [Cursor troubleshooting steps](docs/install/troubleshooting.md#cursor-agent-shell-returns-connect-tunnel-failed-response-403).
  378. - **`calle auth login` failures** — fetch failures, login errors, and token cache issues.
  379. - **Missing legacy MCP tools** — how to confirm that `plan_call`, `run_call`, and `get_call_run` are available for CLI and legacy MCP integrations. For the standalone `calle`, follow its [current tool policy](skills/calle/references/mcp.md#current-call-tools-and-legacy-exclusion).
  380. ## Repository Structure
  381. This is a multi-ecosystem integration monorepo with packaged plugins and standalone Skills.
  382. | Path | Purpose |
  383. | --- | --- |
  384. | `packages/cli` | Shared `calle` CLI for CLI-based integrations. Handles authentication, token caching, MCP tool discovery, and call workflow shortcuts. |
  385. | `packages/core` | Shared core library used across packages. |
  386. | `packages/codex-plugin` | Codex plugin providing the `$calle` skill. |
  387. | `packages/claude-plugin` | Claude Code plugin providing the `/calle:calle` skill. |
  388. | `packages/cursor-plugin` | Cursor plugin bundling the MCP server config, `calle` skill, and real-call safety rule. |
  389. | `packages/openclaw-cli-skill` | OpenClaw CALL-E skill source. |
  390. | `packages/skills-sh-skill` | skills.sh compatible CALL-E skill package. |
  391. | `skills/calle-cli/` | Portable CLI compatibility Skill; official installation uses `--skill calle-cli`. |
  392. | `skills/calle/` | Canonical standalone Skill at version 1.0.6; prefers host-managed OAuth MCP, with direct HTTPS/API Key fallback governed by the [interface selection rules](docs/install/calle.md#choose-the-interface). |
  393. | `tests/` | Offline standalone Skill validation and Agent/mock scenario definitions; execution states are recorded separately. |
  394. | `examples/` | Runnable MCP client demos. |
  395. | `docs/` | Integration guides, install docs, and troubleshooting. |
  396. For layout rules and marketplace naming conventions, see [docs/agent-integration-layout.md](https://github.com/CALLE-AI/call-e-integrations/blob/main/docs/agent-integration-layout.md).
  397. ---
  398. ## Telemetry
  399. The `calle` CLI sends best-effort usage telemetry to help diagnose installation, authentication, and tool availability issues.
  400. **What is collected:** anonymous installation ID, CLI version, integration source (e.g. `claude/claude_code_plugin/<version>`), command stage, outcome, error type, readable `base_url_host` and `server_host` (hostname and any non-default port), and `server_url_hash` (SHA-256 of the complete configured server URL).
  401. The URL hash does not conceal the separate readable host fields. See the
  402. [CLI telemetry field details and example](packages/cli/README.md#telemetry--usage-data).
  403. **What is never collected:** phone numbers, call goals, OAuth tokens, broker login URLs, transcripts, or contact data.
  404. **Opt out** with `--no-telemetry`. First follow
  405. [CLI entry point selection](packages/cli/docs/cli-reference.md#selecting-the-cli-entry-point)
  406. to prepare the launcher and a JSON request. Add `--no-telemetry` to its `argv`:
  407. ```json
  408. ["auth", "status", "--no-telemetry"]
  409. ```
  410. Broker and MCP requests still create service-side security, audit, and operational logs required to run calls.
  411. ---
  412. ## Development
  413. Requires Node `>=22` and pnpm `10.18.3`, Changesets, and GitHub Actions.
  414. ```bash
  415. pnpm install
  416. pnpm check
  417. pnpm test
  418. pnpm pack:dry-run
  419. ```
  420. Standalone Skill checks also require Python 3.10+ and the development dependencies in
  421. `requirements-test.txt`. See [offline validation](tests/README.md) for isolated
  422. setup, test commands, and the separate Agent/mock evaluation status.
  423. <details>
  424. <summary>Package-specific checks</summary>
  425. ```bash
  426. pnpm --filter @call-e/core check
  427. pnpm --filter @call-e/core test
  428. pnpm --filter @call-e/cli check
  429. pnpm --filter @call-e/cli test
  430. pnpm --filter @call-e/codex-plugin check
  431. pnpm --filter @call-e/codex-plugin test
  432. pnpm --filter @call-e/claude-plugin check
  433. pnpm --filter @call-e/claude-plugin test
  434. pnpm --filter @call-e/cursor-plugin check
  435. pnpm --filter @call-e/cursor-plugin test
  436. pnpm --filter @call-e/openclaw-cli-skill check
  437. pnpm --filter @call-e/openclaw-cli-skill test
  438. pnpm --filter @call-e/skills-sh-skill check
  439. pnpm --filter @call-e/skills-sh-skill test
  440. pnpm run check:examples
  441. ```
  442. </details>
  443. For user-visible package changes, add a changeset. The release workflow publishes changed @call-e/* packages to npm and maintains the @call-e/codex-plugin@latest and @call-e/claude-plugin@latest install aliases.
  444. See [CONTRIBUTING.md](https://github.com/CALLE-AI/call-e-integrations/blob/main/CONTRIBUTING.md) for pull request guidelines.
  445. ---
  446. ## Community
  447. - Website: [heycall-e.com](https://www.heycall-e.com/)
  448. - Docs: [docs.heycall-e.com](https://docs.heycall-e.com/)
  449. - Discord: [discord.gg/6AbXUzUV8w](https://discord.gg/6AbXUzUV8w)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.