← ClaudeAtlas

preset-clilisted

Drive Preset's `sup` CLI (PyPI package `superset-sup`) for shell, scripting, CI/CD, and agent-driven Preset workflows. Use only for CLI workflows; Do not use for MCP-only work or for direct HTTP/SDK code paths.
preset-io/agent-skills · ★ 4 · AI & Automation · score 77
Install: claude install-skill preset-io/agent-skills
# preset-cli Use as the foundation for shell, scripting, CI/CD, and agent-driven Preset workflows through the `sup` CLI. ## Always - CLI surface only; stay on MCP or direct API if that's what the user requested. - Default to `--json` for automation and agent consumption. - Keep `SUP_PRESET_API_TOKEN` / `SUP_PRESET_API_SECRET` out of command lines and shared output; use env vars or `sup config auth`, never paste secrets on the command line. - Route push, sync, overwrite, and `--force` to `preset-cli-mutations`. - Redact tokens and credential-bearing output in transcripts. ## Decision Rules - Classify CLI vs MCP vs direct API intent before acting; if MCP or direct API was requested, defer to that plugin. - Distinguish metadata reads from data-returning reads (e.g. `sup sql`, `sup chart data`) — load safety policy before non-metadata reads. - Choose output format based on the downstream consumer: `--json` for automation, `--csv` for files, default Rich for humans, `--porcelain` for shell pipelines. - If a command group is not named on this card, load command coverage before composing commands. - For mutating intent, stop and load `preset-cli-mutations` rather than continuing on this card. ## Workflow Order 1. Establish install, auth, and workspace context. 2. Choose output format. 3. Classify risk (metadata read vs data-returning read vs mutation). 4. Load the focused reference for the operation. 5. Run only the safe / read command. 6. Redact output before sharing. ## Re