菜单
Turn a one-sentence objective into a downloadable .flw workflow that wires the right Tlamatini visual agents and connections.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Run the daily automated Tlamatini chat regression — drive a visible Chrome via Playwright, log into agent_page.html, ask up to 1000 curated safe questions one-by-one (Multi-Turn ON, ACPX/Ask-Execs/Exec-Report/Internet OFF), wait for and qualify each answer (heuristic + LLM judge on failures), then write a dated report + summary. Invoke when the user says "run the daily chat test", "test Tlamatini with the 1000 questions", "daily Tlamatini regression", or schedules this test.
Turn a natural-language objective into a downloadable .flw workflow by driving the FlowCreator engine (full 83-agent catalog), then emit a canvas-loadable .flw.
The authoritative, exhaustive end-to-end runbook for creating a BRAND-NEW Tlamatini workflow agent — every surface, in order, with 530+ numbered steps across 26 phases. Invoke whenever Angela says "create a new agent", "add an agent", "make a <X>er agent", "I want a new canvas agent", or asks to wire any new pool agent across backend + frontend + Multi-Turn + Parametrizer + FlowCreator + FlowHypervisor + watchdog + config dialog + demo prompts + Python tests + Playwright harness tests + docs + packaging. Covers naming, coloring, the inputs/outputs connector contract in agentic_control_panel.html, the Multi-Turn (wrapped chat-agent) tool, Exec Report, the configuration dialog, automated unit tests AND Playwright tests in Claude's harness. Pairs with tlamatini-agent-naming (casing) and the @-imported create_new_agent.md / create_new_mcp.md.
Build and edit in Roblox Studio via the Roblox Studio MCP the RIGHT way - preflight the Studio connection, do the whole build in a few big execute_luau scripts (not dozens of tiny calls), make REALISTIC terrain with the Terrain VOXEL api driven by Perlin noise (NEVER stacked Parts or concentric layers - those give ugly blocky stepped pyramids), poll generative jobs, check the console, and fail honestly. Invoke for ANY "in Roblox / Roblox Studio" request - terrain, mountains, parts, scripts, models, materials, assets.
The authoritative 8-step contract for scaffolding a NEW Tlamatini workflow agent end-to-end (backend script + config.yaml, connection-update view + URL, migration seeding the Agent row, CSS gradient, four JS files, agentic_skill.md, README.md, lint). READ THIS BEFORE adding or renaming any of the 83 visual agents, before touching `agent/agents/<name>/`, before writing a `00NN_add_<name>.py` migration, before adding a `.canvas-item.<x>-agent` rule, or before extending the `acp-*.js` classMap / connectors. Companion to the `tlamatini-new-acp-agent` skill (which drives the procedure) and to `tlamatini-agent-naming` (the naming-convention guard).
The authoritative Tlamatini agent NAMING CONVENTION — invoke before adding, renaming, casing, displaying, or auditing any workflow agent (STM32er, Node Manager, ACPXer, Kalier, ...). Use whenever you touch agentDescription, a pool/agent directory, a CSS .canvas-item class, a JS connection handler, agents_descriptions.md, or any place an agent name is shown. Prevents mis-casing display names (e.g. STM32er must never become STM32Er/STM32ER/Stm32Er).
| name | tlamatini-flow-from-objective |
| description | Turn a one-sentence objective into a downloadable .flw workflow that wires the right Tlamatini visual agents and connections. |
| metadata | {"openclaw":{"emoji":"🌊"},"tlamatini":{"runtime":"in-process","requires_tools":["execute_command","chat_agent_file_creator"],"requires_mcps":[],"budget":{"max_iterations":12,"max_seconds":180,"max_tokens":30000},"permissions":{"filesystem":{"read":["Tlamatini/agent/agents/**/*"],"write":["Tlamatini/**/*.flw"]},"shell":[],"network":"deny","db":"deny"},"inputs":[{"name":"objective","type":"string","required":true,"description":"One-sentence high-level goal"},{"name":"out_path","type":"string","required":true,"description":"Where to write the .flw file. With no user-given folder, default it under the Tlamatini Templates directory (TLAMATINI_TEMPLATES) — never C:\\Temp / %TEMP%."}],"outputs":[{"name":"flw_path","type":"string","required":true},{"name":"agent_count","type":"integer","required":true},{"name":"connection_count","type":"integer","required":true}],"triggers":{"keywords":["flow from","make a flow","build a flow","scaffold flow",".flw"]}}} |
Produce a canvas-loadable .flw for the user's stated objective.
Superseded by the
flow-makingskill. Preferflow-making: it drives the FlowCreator engine (full 83-agent catalog + connection contracts) and emits a validated, schemaVersion-2.flw. This skill is kept as an alias/entry point — do NOT hand-author the.flwJSON, because you do not carry the agent catalog in context and a hand-written flow hallucinates agent types and will not load.
flow-making skill with the same inputs:
invoke_skill('flow-making', { "objective": "${input.objective}", "out_path": "${input.out_path}" }).{ flw_path, agent_count, connection_count }.Use the shipped driver — it copies the FlowCreator template to an isolated dir,
runs it, and writes the .flw:
python Tlamatini/agent/skills_pkg/flow_making/scripts/make_flow.py \
--objective "${input.objective}" --out "${input.out_path}"
The last stdout line is agent_count=<N> connection_count=<M> flw_path=<path>.
.flw shape (schemaVersion 2)If you ever emit .flw JSON by hand, it MUST match the loader contract
(acp-file-io.js::loadDiagram / flow_spec.py) — NOT a {version, agents, connections:[{from,to,kind}]} shape (that is obsolete and will not load):
{
"schemaVersion": 2,
"nodes": [
{"id": "starter-1", "text": "Starter", "left": "50px", "top": "50px",
"agentPurpose": "", "configData": {"target_agents": ["monitor_log_1"]}}
],
"connections": [
{"sourceIndex": 0, "targetIndex": 1, "inputSlot": 0, "outputSlot": 0}
],
"artifacts": {}
}
See agent/skills_pkg/flow_making/references/flw_schema.md for the full contract.