obsidian-ingestlisted
Install: claude install-skill alexeyshishin/as-skill
# obsidian-ingest — Knowledge ingest into the Obsidian vault
This skill is the ingest pipeline coordinator. Your job: take a single markdown file, understand
what knowledge is hidden in it, decide how best to represent it in the base, and create the necessary files.
You do the formatting, frontmatter, and link enrichment yourself — that's an integral part of ingest.
Before starting, read the vault rules:
- `AGENTS.md` — general principles for working with the base
- `.claude/rules/note-types-frontmatter.md` — frontmatter templates and fields
- `.claude/rules/knowledge-structures.md` — atomic notes, MOCs, syntheses
- `.claude/rules/content-style.md` — language, style, wikilinks
- `.claude/rules/tags.md` — tag taxonomy
---
## Step 1. Read and understand the source
Read the input file in full. Determine:
- **Content type**: article / web clip, transcript, meeting notes, draft note, arbitrary markdown
- **Main topic**: one phrase describing what this text is about
- **Domain**: which tags it belongs to (sre, kubernetes, observability, career, concept, etc.)
- **Confidence level**: `medium` if there's a single source, `high` if you see the idea corroborated multiple times
---
## Step 2. Extract entities (knowledge extraction)
Types:
- **Concepts and models** — abstract ideas, frameworks, thinking patterns
- **Insights** — non-trivial conclusions not obvious from the title
- **Facts** — specific statements, data, figures
- **Practices** — "how-to"s, processes, techniques