Setup
Configure credentials, dependencies, and workspace for Look Tongji Notes.
When to Use
- User says
/setup or asks to configure the skill.
- First-time setup before transcribing lectures.
- User wants to change or migrate the course knowledge-base path.
Workflow
- Run setup:
python "<SKILL_DIR>/../../scripts/look_tongji.py" setup
- Non-interactive workspace options:
python "<SKILL_DIR>/../../scripts/look_tongji.py" setup \
--workspace-root "<COURSE_WIKI_ROOT>" \
--owner-name "<OWNER_NAME>" \
--site-name "<OWNER_NAME>的课程知识库"
The CLI checks: Python deps (requests, playwright, python-dotenv, markdown, markitdown), ffmpeg, Node.js, vision-support config, and optionally XeLaTeX.
Vision Support setup: If vision-support/config.json does not exist, the CLI prints the init command. The embedded vision-support is at <SKILL_DIR>/../../vision-support/. Help the user configure a vision provider (OpenAI, Google, Anthropic, deepseek, dashscope, zhipuai, ollama, or custom).
Verify vision-support: After configuration, test with:
node "<SKILL_DIR>/../../vision-support/scripts/vision.mjs" "<SKILL_DIR>/../../komari.jpg"
The agent should correctly identify image content (should mention "red-haired girl" or equivalent).
Never ask for passwords in chat. Use interactive terminal input or environment variables.
Where <SKILL_DIR> Points
<SKILL_DIR> is the directory containing this SKILL.md. Shared scripts (look_tongji.py, timeline_tools.py, tongji_backend/) and references live two levels up in the repository root (<SKILL_DIR>/../../scripts/ and <SKILL_DIR>/../../references/).
1---2name: setup3description: Configure Tongji Look credentials, check system dependencies (Python, Node.js, ffmpeg, vision-support, XeLaTeX), and set up the persistent course-wiki workspace.4license: MIT5---67# Setup89Configure credentials, dependencies, and workspace for Look Tongji Notes.1011## When to Use1213- User says `/setup` or asks to configure the skill.14- First-time setup before transcribing lectures.15- User wants to change or migrate the course knowledge-base path.1617## Workflow18191. Run setup:2021```bash22python "<SKILL_DIR>/../../scripts/look_tongji.py" setup23```24252. Non-interactive workspace options:2627```bash28python "<SKILL_DIR>/../../scripts/look_tongji.py" setup \29 --workspace-root "<COURSE_WIKI_ROOT>" \30 --owner-name "<OWNER_NAME>" \31 --site-name "<OWNER_NAME>的课程知识库"32```33343. The CLI checks: Python deps (requests, playwright, python-dotenv, markdown, markitdown), ffmpeg, Node.js, vision-support config, and optionally XeLaTeX.35364. **Vision Support setup:** If `vision-support/config.json` does not exist, the CLI prints the init command. The embedded vision-support is at `<SKILL_DIR>/../../vision-support/`. Help the user configure a vision provider (OpenAI, Google, Anthropic, deepseek, dashscope, zhipuai, ollama, or custom).37385. **Verify vision-support:** After configuration, test with:39 ```bash40 node "<SKILL_DIR>/../../vision-support/scripts/vision.mjs" "<SKILL_DIR>/../../komari.jpg"41 ```42 The agent should correctly identify image content (should mention "red-haired girl" or equivalent).43446. Never ask for passwords in chat. Use interactive terminal input or environment variables.4546## Where `<SKILL_DIR>` Points4748`<SKILL_DIR>` is the directory containing this `SKILL.md`. Shared scripts (`look_tongji.py`, `timeline_tools.py`, `tongji_backend/`) and references live two levels up in the repository root (`<SKILL_DIR>/../../scripts/` and `<SKILL_DIR>/../../references/`).
WALKERKILLER