Arnav

Sets up a structured wiki in Obsidian that Claude can search, cross-reference, and synthesize—so your team can ask questions about accumulated knowledge instead of scrolling.

Takes a file, URL, or folder and breaks it into interconnected wiki pages in Obsidian—each a focused concept with backlinks to related ideas.

Queries your Obsidian vault for everything related to a topic and synthesizes answers with exact links back to the original notes.

Creates folder structure, templates, config files, and .env setup for a personal wiki. Saves hours of scaffolding so you start writing notes immediately.

Scans an Obsidian vault for orphaned pages, broken links, stale notes, and contradictions. Outputs a repair checklist ranked by impact.

Reports which docs have been ingested, which are queued, and what changed since the last sync — so you know what's live versus what's still pending.

Extracts the key decisions, learnings, and artifacts from your project and pushes a formatted note into your Obsidian vault without overwriting existing entries.

Snapshots all current wiki pages, then rebuilds from scratch using fresh sources—Slack threads, docs, repos—so outdated info doesn't resurface.

Scans your Obsidian vault and surfaces which pages should link to each other based on semantic relevance. Suggests the link text and anchor so you just approve.

Imports a conversation thread or session log and turns it into a wiki—indexed by topic, decision, and outcome so you can search months of chats later.

Reads your Obsidian vault and outputs the note structure as JSON, GraphML, or Neo4j format—so you can visualize connections, query patterns, or feed it into other tools.

Scans your wiki, finds tag drift and inconsistency, and rewrites pages to use a single controlled vocabulary — so searches actually work.

Ingest Claude Code conversation history into the Obsidian wiki. Use this skill when the user wants to mine their past Claude conversations for knowledge, import their ~/.claude folder, extract insights from previous coding sessions, or says things like "process my Claude history…

Reads unstructured conversation exports (Slack, Discord, meeting transcripts) and structures them into an Obsidian wiki with linked notes, timestamps, and full-text search.

Scaffolds a new skill from scratch, runs evals against test cases, measures performance with variance analysis, and identifies optimization opportunities without trial-and-error.

Reads your Obsidian vault and rewrites the graph-view config to color-code nodes by tag, folder, or custom category — making visual clusters pop.

Runs multi-round web searches on a topic, synthesizes findings into a structured outline, and writes it directly into your Obsidian vault with linked references.

Builds a dynamic table or index view inside Obsidian that surfaces all notes matching a tag, type, or date—updated live as you add content.

Extracts the key ideas, decisions, and context from your current chat and writes them into a structured, searchable wiki entry you can reference later.

Scans your Obsidian vault for concepts that appear together on multiple pages but lack a synthesis page, then drafts connective pages that draw cross-cutting conclusions.

Reads your ~/.codex folder, mines past coding sessions for reusable patterns, and builds an Obsidian wiki you can search and link through later.

Fetches any URL, extracts the core content, and files it into your Obsidian vault—either in a project folder or a catch-all inbox that auto-organizes as you link notes.

Reads your local Hermes conversation log, extracts insights and decisions, and populates an Obsidian vault with linked notes—so you can search past work and recall what you learned.

Pulls targeted transcripts from a specific agent's past conversations—Claude, Codex, Hermes, or others—filtered by topic or timeframe to find prior work or patterns.

Runs your daily maintenance cycle: validates that all source docs are current, refreshes the index, and regenerates your hot-items list automatically.

Searches a shared wiki across Claude, Codex, Hermes, and other tools to show you which gaps exist and where one tool has insight the others don't.

Reads your OpenClaw conversation history and imports it into Obsidian, tagged and linked so you can search past reasoning, find related decisions, and spot patterns across projects.

Takes a stated goal and what you built, then gives honest feedback: Does it match? What's missing? Where would it break in real use?

Switches your Obsidian context between work, personal, project, or archive vaults without closing the app. See which vault you're in and add new profiles on the fly.

Scans your wiki for new, updated, and linked pages over the past week and writes a newsletter-style summary showing patterns and connections without overwhelming detail.

Ingests a batch of docs into Obsidian, dedupes against what's already there, auto-generates frontmatter and cross-links so notes stay searchable and connected.

Reads your Copilot CLI history and extracts the architecture decisions, debug patterns, and code insights into timestamped, linked Obsidian pages you can search and reference later.

Scans your vault for pages covering the same concept under different names—like "RSC" and "React Server Components"—and proposes smart merges with redirects intact.

Finds the most relevant pages in your Obsidian vault on a topic and bundles them into a compact, structured summary that fits a Claude context window — ready to hand off to another skill.

Displays pending wiki edits side-by-side with live versions, lets you accept or reject each change, and publishes approved ones in bulk without manual file moves.

Imports your Pi agent session history into an Obsidian wiki, making past coding decisions, solutions, and debugging steps searchable and reusable across projects.

Takes a graph.json export and reconstructs the full wiki structure—pages, links, metadata—into your current vault without losing hierarchy or connections.

Pulls the insight or fix from your current chat thread and drops it into your wiki's staging area—no formatting, no friction, ready to polish later.

>