# Cuelist > Import a song's cue-list CSV (section markers like PGM IN / INTRO / VER 1 / CHOR 1 with a number, name, and timecode) into that song's grandMA2 sequence, building each section as a labelled cue so the operator gets a navigable skeleton to program against. Use whenever the user hands over a per-song CSV of cues/sections and wants it pushed onto the console — "save this cue list to the sequence", "import these cues", "build the cues from this CSV", "turn this section map into cues for the song", or points at a file named like `S1_.csv`. Assumes the song's sequence already exists (built by the setlist skill) and the gma2 MCP is connected (connect skill). Prefer this over hand-typing Store commands — it parses the CSV, maps it to the right sequence, guards against clobbering existing cues, and verifies the result. - Skill: `chienchuanw/cuelist` (Agent Skill, multi-file: 4 files) - Install (CLI): `npx skillmds@latest add chienchuanw/cuelist` - Raw SKILL.md: https://api.skillmd.com/api/skills/chienchuanw/cuelist/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Data & Analytics - Author: chienchuanw (https://skillmd.com/u/chienchuanw) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/chienchuanw/cuelist --- # gma2 cuelist Take a per-song cue-list CSV and store each row as a labelled cue inside that song's sequence. The cues are **empty placeholders** — a navigation skeleton (PGM IN → INTRO → VER 1 → …) that the operator walks with Go and programs later. The CSV is the section map; the lighting comes afterwards. Parsing lives in `scripts/csv_to_cues.py` (a pure transform — it only prints commands, never touches the desk). All console writes go through the gma2 MCP `send_raw_command`, so the console path stays in one place. ## Inputs (from the user) - **The CSV** — one row per musical section. Columns vary by export tool, but there's always a **name** (PGM IN, INTRO, VER 1, CHOR 1, FILL, RISE, END…), usually a **cue number** and a **time** (seconds into the song), and often `fadeIn`/`fadeOut`/`notes`. Files are commonly named `S_.csv` where `` is the song's order within the show. - The song's **sequence already exists** (built by the `setlist` skill, with a home Cue 0.5 "LOADING SONG"); the console is connected (`connect` skill). ## Workflow ### 1. Inspect the CSV Run the parser to see the sections and catch oddities before writing anything: ```bash python3 plugins/gma2/skills/cuelist/scripts/csv_to_cues.py ".csv" --dump ``` It prints a `cue · name · time` table, the time span, and warnings (blank names, duplicate names — legal since cue numbers differ, non-increasing numbers). Column matching is forgiving (case-insensitive keywords), so most exports parse without fuss. Report the table so the user can eyeball it. ### 2. Resolve the target sequence (propose, then confirm) The CSV filename usually carries the **song title** (often in the user's language), while the sequence is labelled with the one-word English name the setlist skill assigned. These don't map mechanically, so **don't guess silently** — read the live sequences and propose the match: ``` List Sequence # via send_raw_command — lists song sequences by name ``` Use the `S` order in the filename plus the song-title meaning to propose the matching sequence (e.g. `S1_食等睡等` → the first song → Sequence 101 "Idle"). State your reasoning and **confirm the sequence number with the user** before writing. Getting this wrong writes cues onto the wrong song. ### 3. Guard against clobbering existing cues A freshly-built song sequence holds only the home Cue 0.5, so `List Sequence ` reports `(1)` at the end of its row — the **cue count**. Read it first: ``` List Sequence # row ends with (N) = number of cues ``` If `N > 1`, the sequence already has show cues — **stop and ask** before overwriting or extending. Don't silently store over someone's programming. (A count of exactly 1 means only the home cue exists → safe to add 1..N.) ### 4. Build the cues Generate the command list for the confirmed sequence and send each line through `send_raw_command`: ```bash python3 .../csv_to_cues.py ".csv" --commands ``` This emits `ClearAll`, then one `Store Sequence Cue "" /nc` per row (label baked into the Store), then a closing `ClearAll`. The `ClearAll` brackets matter: storing from an empty programmer is what makes each cue an **empty, labelled placeholder** rather than capturing whatever was live. The home Cue 0.5 is left untouched. **Defaults and why:** - **Empty cues, manual Go trigger** — the setlist engine and its Prev/Next/Goto buttons are a hands-on operating model; the cues are a structure to program into, not an auto-playing timeline. The timecode column is kept as *reference*, not written, by default. - If the user wants the times recorded on the cues, add `--note-col time` to write `@s` into each cue's Note (a separate `Assign Cue … /note` line). For auto-advancing cues (Follow / Time / Timecode) see `references/cue-options.md` — that's a deliberate opt-in, not the default. ### 5. Verify and report Read the cues back: ``` List Cue Sequence ``` Confirm the count went from `(1)` to `(1 + rows)` and that the new cues show `Trig = Go`, `Mode = Normal`. **Cue Name is not exposed over this console's Telnet** (a known limitation — see the connect skill), so the labels can't be read back; confirm them by eye on the console. Report the cue table (number · name · time), note the duplicate-name flags, and remind the user the show isn't saved (Backup / Save Show). ## Notes - The parser never touches the console — it only prints commands. Every write goes through the MCP, keeping one uniform console path (same pattern as the other gma2 skills). - Cue numbers come from the CSV's number column (integers stay integers); if there's no number column it falls back to row order 1..N. - One CSV = one song = one sequence. To do a whole show, run this per song.