migrate-from-elevenlabs
Skill0 starsChanged 58 days agoMITView on GitHub
Import and validate ElevenLabs conversational agents in ThunderPhone. Use when someone asks to migrate from ElevenLabs Agents, create an ElevenLabs import job, review voice or knowledge mappings, identify client-tool or MCP gaps, commit selected items, or run side-by-side calls before cutover.
- Reads credentials
- Sends data out
In this repository21 other files it gives its agents.
- AGENTS.mdAGENTS.md
- add-knowledgeSkill
- build-tool-integrationSkill
- create-agentSkill
- embed-web-widgetSkill
- get-phone-numberSkill
- handle-inbound-callsSkill
- iterate-on-agentSkill
- migrate-from-blandSkill
- migrate-from-retellSkill
- migrate-from-vapiSkill
- place-outbound-callsSkill
- run-campaignSkill
- setup-api-keySkill
- setup-webhooksSkill
- test-agentSkill
- thunderphone-mcp-setupSkill
- thunderphone-prompt-builderSkill
- use-with-livekitSkill
- use-with-openai-realtimeSkill
- use-with-pipecatSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Choose import mode: Use authorized api_key discovery or a permitted paste payload.
- Distinctive: line 66
Run side-by-side tests: Use identical scenarios for source and ThunderPhone.
- Distinctive: line 85
Use import agents and the agent-import API.
- Distinctive: line 86
Never expose source secrets or assume a source voice exists in ThunderPhone.
- Distinctive: line 87
Keep source agents and traffic intact until matched tests and rollback are complete.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Migrate from ElevenLabs
- Inventory the source. Export conversational agents, prompts, first messages, voices, LLM settings, tools, knowledge, numbers, secrets, traffic, and test cases. Record the source version and content owners.
- Choose import mode. Use authorized
api_keydiscovery or a permittedpastepayload. Keep the source credential in the protected API flow and out of prompts, JSON files, output, and logs. - Create the job at
POST /v1/agent-importswithsource: elevenlabs.
{
"source": "elevenlabs",
"auth_mode": "paste",
"payload": "${EXPORTED_AGENT_JSON}"
}curl --fail-with-body -X POST https://api.thunderphone.com/v1/agent-imports \
-H "Authorization: Bearer $THUNDERPHONE_API_KEY" -H "Content-Type: application/json" \
-H "X-ThunderPhone-Client: skills/migrate-from-elevenlabs@1.0" --data-binary @elevenlabs-import.jsonimport json, os, requests
with open("elevenlabs-import.json", encoding="utf-8") as f: payload = json.load(f)
r = requests.post("https://api.thunderphone.com/v1/agent-imports", json=payload,
headers={"Authorization": f"Bearer {os.environ['THUNDERPHONE_API_KEY']}",
"X-ThunderPhone-Client": "skills/migrate-from-elevenlabs@1.0"}, timeout=60)
r.raise_for_status(); print(r.json())import { readFile } from "node:fs/promises";
const r = await fetch("https://api.thunderphone.com/v1/agent-imports", { method: "POST",
headers: { Authorization: `Bearer ${process.env.THUNDERPHONE_API_KEY}`,
"Content-Type": "application/json",
"X-ThunderPhone-Client": "skills/migrate-from-elevenlabs@1.0" },
body: await readFile("elevenlabs-import.json", "utf8") });
if (!r.ok) throw new Error(await r.text()); console.log(await r.json());- Poll and review items. Fetch
GET /v1/agent-imports/{public_id}until ready; usePATCH /v1/agent-imports/{public_id}/items/{item_id}to select and correct each candidate. - Review mapping. Prompt, first message, language, LLM choice, voice, webhook tools, supported system tools, knowledge, silence, and maximum duration can map. Numbers, secret values, client-side tools, MCP servers, and unsupported vendor capabilities need manual work. See references/elevenlabs-mapping.md.
- Commit only reviewed items with
POST /v1/agent-imports/{public_id}/commit; poll to terminal and record each created ID. Discard an uncommitted job withDELETE /v1/agent-imports/{public_id}. - Run side-by-side tests. Use identical scenarios for source and ThunderPhone. Compare tool and knowledge evidence, transfer/end behavior, language, pronunciation, interruption, audio, and latency. Create a separate approved plan for numbers, traffic, and source retirement.
- Read current comparison data from
https://thunderphone.com/data/voice-ai-price-index.json; do not publish invented or stale rates.
Errors
| Symptom | Cause | Fix |
| Voice does not map | Source voice is unavailable or incompatible | Select a real voice returned by GET /v1/voices and evaluate it in audio. |
| Client tool disappears | Browser/client-executed tools are not server tools | Rebuild an authorized server integration or keep the client layer explicitly. |
| Knowledge answers differ | Source documents or retrieval behavior did not transfer fully | Import permitted sources, run direct search, then test spoken answers. |
| MCP capability is missing | Remote MCP servers are manual assets | Register them through /v1/mcp-servers, review discovered tools, and attach deliberately. |
Source and safety rules
- Use import agents and the agent-import API.
- Never expose source secrets or assume a source voice exists in ThunderPhone.
- Keep source agents and traffic intact until matched tests and rollback are complete.
- Separate transcript similarity from audio, latency, connection, and tool evidence.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

