The maintained Obsidian AI Skills Directory is the canonical place to browse available skills, source repositories, compatible agents, and individual skill pages. This page explains how to choose among them without treating a long list of repository names as a recommendation.
This page is intentionally not an exhaustive ranking. AI skills change as their source repositories, agent formats, and Obsidian features change. A skill that is useful for a terminal-based workflow may be unnecessary for a chat plugin, and a skill that writes valid Markdown may still be a poor fit for a vault with sensitive notes.
Start With the Workflow, Not the Skill Name
Choose the outcome you want first:
Use the directory filters to narrow the candidates, then open the individual resource page and source repository. The directory is for discovery; the repository and the skill's own documentation are the authority for current behavior.
Start With These Skill Pages
These cards are a curated starting set, not a second copy of the full catalog. Open a card to read its installation notes, compatibility details, sources, and usage guidance.
Obsidian Markdown
Teaches AI agents how to use Obsidian-flavored Markdown syntax correctly.
- Giving AI agents a consistent Obsidian Markdown syntax reference
- Preventing agents from writing standard Markdown when Obsidian syntax is required
- Mastering callout types, math blocks, and internal link aliases
Obsidian Bases
Allows AI agents to manage Obsidian Bases (.base files), formulas, and database views.
- Building structured note databases in Obsidian
- Defining property filters and formulas in .base files
- Automating table views across vault categories
Obsidian CLI
Teaches AI agents how to interact with Obsidian vaults via command-line interface.
- Automating vault tasks from the terminal
- Integrating AI agent workflows with local Obsidian apps
- Triggering plugin commands programmatically
Obsidian Wiki Suite
One canonical guide for ingesting material, querying linked notes, and updating an Obsidian wiki.
- Testing a source-to-Markdown knowledge workflow in a disposable vault
- Querying a linked note collection while preserving note references
- Reviewing incremental wiki updates instead of rebuilding a corpus blindly
Obsidian Second Brain Workflow
Open-source second brain workflow for organizing notes, knowledge, and personal learning.
- Setting up a personal second brain in Obsidian
- Organizing research and learning materials
- Standardizing note links and index pages
A Simple Selection Method
Evaluate no more than two or three candidates for the same job. For each candidate, record:
- Input: What files, prompts, commands, or external sources does it read?
- Output: What does it create or change in the vault?
- Control: Can you preview, approve, undo, or limit its changes?
- Maintenance: Who maintains it, when was it last updated, and which agent formats does it support?
- Data flow: Can it send note content to a model, website, API, or third-party integration?
Prefer the smallest skill that solves the immediate task. Combining several overlapping instruction packages can produce conflicting folder rules, duplicated notes, or unexpected edits. Add another skill only after the first one has a clear boundary and a tested result.
Test Any Skill on a Copy of the Vault
Before installing a skill into an irreplaceable vault, make a small test vault containing representative files:
- one ordinary Markdown note;
- one note with YAML properties;
- one note with wikilinks and a callout;
- one attachment or embedded file;
- one intentionally malformed example to test error handling.
Run one read-only request first. Check whether the agent cites the files it used and whether its proposed output matches your naming and folder rules. Only then allow a write operation, and review the diff or changed files immediately afterward. Keep a backup before any batch command.
Security and Privacy Checks
Treat a skill as executable instructions, not as harmless prompt text. Read its source, license, installation steps, and update history. Check whether it asks an agent to run shell commands, browse the web, install dependencies, call an API, or transmit vault content. Keep sensitive notes outside the test vault until the data flow is clear.
For terminal agents, start from a narrowly scoped folder or use explicit paths. For community plugins, review the plugin's permissions and repository before enabling it. A skill can improve an agent's Obsidian formatting while still having broad access to the files available to that agent.
Where to Continue
- Browse the current Obsidian AI Skills Directory.
- Read the Obsidian Markdown skill when formatting correctness is the first requirement.
- Follow the Claude Code and Obsidian tutorial for a controlled terminal workflow.
- See the Obsidian second-brain workflow for a review-first system that separates durable notes from generated drafts.
The right choice is the skill with the clearest input, output, permissions, and maintenance boundary for your vault. A shorter, well-tested workflow is more valuable than installing every skill in a directory.
