# Lisa Add Ingest

> Scaffold a project-specific front-door ingest skill that does something unique (classify a source, fetch from a special system, stamp domain frontmatter) and then chains into /ingest. Extends ingestion without forking the kernel.

- Skill: `codyswanngt/lisa-add-ingest-2` (Agent Skill)
- Install (CLI): `npx skillmds@latest add codyswanngt/lisa-add-ingest-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/codyswanngt/lisa-add-ingest-2/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: codyswanngt (https://skillmd.com/u/codyswanngt)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/codyswanngt/lisa-add-ingest-2

---

## Lisa Command Compatibility

- Original Claude command: `/lisa:add-ingest`
- OpenCode invocation: `$lisa-add-ingest` or a plain-English request that matches this skill.
- Treat the user's surrounding request as the command arguments.
- Claude argument hint: `<short name for the new ingest path>`

Use the lisa-wiki-add-ingest skill to generate a thin, registered front-door ingest skill: interview the project for the source type, bucket, frontmatter, and side-effect class; emit a lisa-wiki-local-<name> skill that enriches then chains into /ingest; and register it under customConnectors. Use the user's surrounding request as this command's arguments.