Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Obsidian CLI · kepano | DecimalAI

Obsidian CLI

SkillScore42Maintained

obsidian-cli

Interact with Obsidian vaults using the Obsidian CLI to read, create, search, and manage notes, tasks, properties, and more. Also supports plugin and theme development with commands to reload plugins, run JavaScript, capture errors, take screenshots, and inspect the DOM. Use when the user asks to interact with their Obsidian vault, manage notes, search vault content, perform vault operations from the command line, or develop and debug Obsidian plugins and themes.

Documentationby kepano

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Knowledge Management pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/obsidian-cli/SKILL.md

decimalai skills pull obsidian-cli --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/obsidian-cli/SKILL.md https://app.decimal.ai/s/obsidian-cli/SKILL.md

Raw file: /s/obsidian-cli/SKILL.md ↗ · machine-readable summary at /s/obsidian-cli.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

31 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 67d ago

Yes — +39% lift-9% tokens0% turns42% of whole cases pass with skill (13/31)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-17	✗→✓	▲ Improved	—	—
case-19	✗→✓	▲ Improved	—	—
case-02	✗→✓	▲ Improved	—	—
case-26	✗→✓	▲ Improved	—	—
case-22	✗→✓	▲ Improved	—	—

Showing 5 of 31 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 7/30/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

kepano / obsidian-skills

skills/obsidian-cli/SKILL.md

View on GitHub ↗

Imported 6/14/2026Version v1

Skill contents

SKILL.md

Obsidian CLI

Use the obsidian CLI to interact with a running Obsidian instance. Requires Obsidian to be open.

Command reference

Run obsidian help to see all available commands. This is always up to date. Full docs: https://help.obsidian.md/cli

Syntax

Parameters take a value with =. Quote values with spaces:

bash
obsidian create name="My Note" content="Hello world"

Flags are boolean switches with no value:

bash
obsidian create name="My Note" silent overwrite

For multiline content use \n for newline and \t for tab.

File targeting

Many commands accept file or path to target a file. Without either, the active file is used.

file=<name> — resolves like a wikilink (name only, no path or extension needed)
path=<path> — exact path from vault root, e.g. folder/note.md

Vault targeting

Commands target the most recently focused vault by default. Use vault=<name> as the first parameter to target a specific vault:

bash
obsidian vault="My Vault" search query="test"

Common patterns

bash
obsidian read file="My Note"
obsidian create name="New Note" content="# Hello" template="Template" silent
obsidian append file="My Note" content="New line"
obsidian search query="search term" limit=10
obsidian daily:read
obsidian daily:append content="- [ ] New task"
obsidian property:set name="status" value="done" file="My Note"
obsidian tasks daily todo
obsidian tags sort=count counts
obsidian backlinks file="My Note"

Use --copy on any command to copy output to clipboard. Use silent to prevent files from opening. Use total on list commands to get a count.

Plugin development

Develop/test cycle

After making code changes to a plugin or theme, follow this workflow:

Reload the plugin to pick up changes:

bash obsidian plugin:reload id=my-plugin

Check for errors — if errors appear, fix and repeat from step 1:

bash obsidian dev:errors

Verify visually with a screenshot or DOM inspection:

bash obsidian dev:screenshot path=screenshot.png obsidian dev:dom selector=".workspace-leaf" text

Check console output for warnings or unexpected logs:

bash obsidian dev:console level=error

Additional developer commands

Run JavaScript in the app context:

bash
obsidian eval code="app.vault.getFiles().length"

Inspect CSS values:

bash
obsidian dev:css selector=".workspace-leaf" prop=background-color

Toggle mobile emulation:

bash
obsidian dev:mobile on

Run obsidian help to see additional developer commands including CDP and debugger controls.

Illustrative Want to see what obsidian-cli would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

7/23/2026, 06:26 UTC · 31 cases

Lift

+39%

42% vs 3%

Agent turns

1.0 vs 1.0

Tokens

-9%

1,319.59 vs 1,451.22

fail → pass · 12pass → pass · 1fail → fail · 18

Per-case results (31)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-04 Can you tell me how many notes I have under each tag, ranked from most used to least, give me just the final number rather than a big listing, and put the whole thing somewhere I can paste into another document without manually copying it out of a window myself?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-23 I want to check what a specific CSS property currently resolves to on a particular element in the live app — say, its background color — without opening dev tools by hand. What's the shape of the command, and which two things does it need me to specify?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-17 Whatever command I end up running — a read, a search, a listing, anything — I want its output copied straight to my clipboard instead of just printed to the terminal. Is that tied to one specific command, or can it be added generally? What exactly do I add?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-07 I want to append two new lines to the end of an existing note in a single call — not just one line, and not two separate append operations. When I build the text string I hand to the append command, what exact character sequence do I put in it to represent the break between the two lines?	pass→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-10 I've only got one Obsidian window open right now, and I run a command without specifying which vault to use at all. Which vault does the command act against by default?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-31 I don't have any vault set up yet for a new project — I want a brand-new vault created from scratch, called Research, with a note dropped into it once it exists. Can the same targeting mechanism that points a command at a particular existing vault also spin up that new vault for me first?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-19 For a command that would normally print a full listing of items, I just want the count of how many there are, not the actual list itself. What do I add to switch it into count-only mode?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-02 I've got both my Work vault and my Personal vault open in Obsidian right now. There's a note in the Personal one — I only remember its title, not which folder it's tucked into — can you pull up everything that links back to it, making sure you're definitely looking inside the Personal vault and not whichever one happens to be active at the moment?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-26 Beyond the screenshot, DOM, console, and CSS inspection commands, I've heard there are even lower- level developer commands available — things like remote-debugging protocol access or debugger controls. Where would I find the full list of those, since they aren't all spelled out to me already?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-22 I want to run one quick piece of JavaScript directly inside the live app's own runtime — say, counting how many files are in the vault — rather than writing a whole plugin just for that. What's the shape of the command for that, and where does the code itself go in it?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-29 I want you to start reading and searching through my notes right now, but I haven't actually opened the Obsidian app itself yet today — is that something this tool can do without me opening it first, or does the app need to already be running for any of this to work?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-03 Create a new note capturing today's meeting. It needs three distinct lines — an agenda line, then a decisions line nested underneath it, then a follow-ups line — all set up in one shot so I don't have to go back and reformat it, and I don't want it popping open on screen when you're done.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-09 If I run a command that normally needs to know which note to act on, but I supply neither of the two note-targeting parameters at all, what does the command fall back to acting on?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-30 I found a plugin in the community plugin browser that I haven't installed yet, and I'd like it fully installed and enabled in my vault. Can the same tool that handles picking up code changes for plugins I've already got installed also handle grabbing and installing this brand-new one from the store for me?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-15 I want to set a frontmatter property called status to the value done on one specific existing note. What's the shape of the command for that — name the parameters it needs and what each one carries.	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-21 I just reloaded a plugin after a code change, and the error-check step comes back showing an actual error. Per the documented workflow, what's the very next thing I should do — proceed on to the visual-verification step anyway, or something else?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-28 Walk me through the documented order of steps for verifying a plugin change end-to-end in this workflow — from picking up the code change through confirming it visually and checking for suspicious log output. List the steps in order, not just describe them.	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-05 I'm about to run a command against this tool and one of my values has spaces in it — a note title like "Weekly Sync Notes." In general terms, how does this CLI expect a parameter's value to be attached to its name, and what do I need to do differently for a value with spaces versus one without?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-27 I have an existing note and I want to add one new line onto the end of it — I do NOT want a brand new note created, and I do NOT want the existing content touched otherwise. Which of the two note-writing commands is the right one for that, versus the one meant for making an entirely new note?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-16 I want a single command that shows just today's still-open to-do items, scoped to the daily note only — not every task across my whole vault. What's that command?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-20 I want the single most authoritative, always-current place to look up this CLI's full command list and syntax, rather than relying on a doc that might be stale. What do I run to get that?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-14 I want two separate things: first, just print today's daily note as it currently stands, and second — as a separate call — add a new checklist item onto the end of it. Give me the shape of each of those two commands.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-18 Separately from note-creation, I'm running a property update on an existing note and I don't want that note's file to pop open in the app as a side effect. What do I add to suppress that, and is this flag specific to note-creation or usable more broadly across commands?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-13 I want to search my vault for a term and cap the number of results returned so I don't get flooded. What's the shape of that command — which parameter carries the search term, and which one caps the result count?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-12 I want to create a brand-new note that's built from one of my existing templates, and I also don't want it popping open once it's made. What does the command for that need to include, at minimum — name the pieces.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-25 Rather than a screenshot, I want the actual text content currently rendered inside a specific part of the interface, identified by its selector, printed back to me directly. What's the shape of the command for that kind of DOM inspection, and what do I add to get text back instead of structure?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-24 I want to preview how the app would render on a phone-sized screen without actually switching devices — some kind of emulation toggle. What's the shape of the command that turns that on?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-06 Separately from parameters that take a value, this tool also has switches that just turn a behavior on with no value attached — say, something that stops a note from opening, or something that allows overwriting an existing item. How are those switches written on the command line compared to a parameter=value pair, and can more than one be added to the same command at once?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-11 Give me the general shape of the command I'd run to just print the full text of one existing note, when all I have is its title.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-08 This tool lets me point a command at a specific note in two different ways: one where I supply something like the note's title, and one where I supply its exact location inside the vault. Explain the difference between those two targeting styles — what does each one expect me to supply, and does the title-based style need me to include a folder or a file extension?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-01 I edited some code in one of my Obsidian plugins. I need the update running in the app, confirmation nothing broke, visual proof it's actually rendering correctly, and a heads up on anything suspicious that got logged — handle whatever's needed if something still looks off before you call it done.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 31 cases were attempted, and 27 counted toward the lift figure. The other 4 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +39 percentage points is the difference between those two pass rates over the 27 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: no judge model recorded
Grading method: judged
Verification: verified
Pass rate without the skill: 3.2%
Pass rate with the skill: 41.9%
Cases attempted: 31
Cases counted toward lift: 27
Cases that regressed: 2
Skill version measured: v1
Run completed: 7/23/2026, 06:26 UTC

The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run42

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns-3% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 2 months ago

Other measured skills in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

Obsidian CLI

SkillSafety

Source

Obsidian CLI

Command reference

Syntax

File targeting

Vault targeting

Common patterns

Plugin development

Develop/test cycle

Additional developer commands

How this skill was measured

Related skills