ar9av/obsidian-wiki
35 skills78.3k installs109k starsGitHub
Install
npx skills add https://github.com/ar9av/obsidian-wikiSkills in this repo
1Llm Wikillm-wiki documents the foundational knowledge distillation pattern for building and maintaining an AI-powered Obsidian wiki based on Andrej Karpathy's LLM Wiki architecture. Use it when users want to understand wiki theory, set up a knowledge base, or decide how raw sources compile into interconnected markdown pages. The three layers are immutable raw sources in OBSIDIAN_SOURCES_DIR, an LLM-maintained wiki vault at OBSIDIAN_VAULT_PATH with YAML frontmatter and wikilinks, and a schema governed by categories, page templates, and operational workflows. Default categories include concepts, entities, skills, references, synthesis, and journal, with optional project-scoped directories under projects/. Special files index.md, log.md, and .manifest.json enable delta ingest, provenance audit, and append versus rebuild decisions. Pages carry provenance markers for extracted, inferred, and ambiguous claims, typed relationships frontmatter, base_confidence and lifecycle states, and importance tiering for ingest and query prioritization. Config resolves by walking up for .env with OBSIDIAN_VAULT_PATH or falling back to ~/.obsidian-wiki/config. Companion skills handle ingest, query, lint, statu.3.4kinstalls2Wiki Ingestwiki-ingest is an Obsidian wiki skill that distills and integrates knowledge from heterogeneous sources into linked vault pages rather than producing isolated summaries. It resolves vault config from .env or ~/.obsidian-wiki/config, honors WIKI_STAGED_WRITES by landing pages in _staging for review, and reads .manifest.json plus index.md and log.md before processing. Three ingest modes cover append with content-hash deduplication, full re-ingest, and raw promotion from _raw/ drafts. Supported sources include markdown, text, PDF with vision for academic figures, JSON and CSV, chat exports, images, folders, and web URLs via references/url-sources.md. A content trust boundary treats source files as untrusted data and forbids executing embedded instructions. The process extracts concepts, entities, procedures, attributable claims, and cross-links while noting contradictions with existing wiki content. Append mode skips unchanged files even when mtimes drift by comparing SHA-256 hashes against the manifest. Developers reach for wiki-ingest when research, meeting notes, chat logs, or URLs should become durable second-brain pages.3.2kinstalls3Wiki SetupThe wiki-setup skill initializes a new Obsidian wiki vault with the correct structure, special files, and configuration. Use it when setting up a wiki from scratch, creating a new vault, fixing a broken setup, or reconfiguring an existing vault. Step one creates a .env from .env.example asking for OBSIDIAN_VAULT_PATH, OBSIDIAN_SOURCES_DIR, optional CLAUDE_HISTORY_PATH, QMD collection settings, WIKI_TOKEN_WARN_THRESHOLD, and WIKI_STAGED_WRITES for review-before-promote workflows. Step two creates directories: concepts, entities, skills, references, synthesis, journal, projects, _archives, _raw, _staging, and .obsidian. Step three seeds index.md, log.md, and hot.md with frontmatter templates. Step four writes minimal .obsidian app.json and appearance.json for preview mode. Step five recommends Dataview, Graph Analysis, Templater, and Obsidian Git plugins. Step six verifies vault paths, special files, and source readability. Optional stop-hook installation auto-captures session findings via wiki-capture. After setup users can run wiki-status, wiki-ingest, claude-history-ingest, and codex-history-ingest. QMD update runs when semantic search collections are configured.3.1kinstalls4Wiki Querywiki-query answers questions against a compiled Obsidian knowledge base using a tiered retrieval protocol defined alongside llm-wiki. It is read-only except for a single append to log.md and routes write requests to wiki-capture or wiki-update skills. Config resolution walks CWD for .env, then ~/.obsidian-wiki/config, yielding OBSIDIAN_VAULT_PATH and optional QMD collection variables. Retrieval starts with hot.md and index.md, builds frontmatter-only candidates via grep, optionally runs QMD semantic or lex vec queries when QMD_WIKI_COLLECTION is set, escalates to section grep, then full reads of up to three tier-ordered pages. Index-only mode triggered by quick answer or fast lookup phrases answers from summaries without reading bodies. Multi-hop path queries walk relationships frontmatter with bounded BFS depth three and frontier cap near sixty pages, reporting typed edges like extends, contradicts, and derived_from. Visibility filtered mode excludes visibility/internal and visibility/pii tagged pages when users request public-only answers. Answers cite wiki pages, annotate lifecycle staleness or disputed status, and propose rather than perform content changes.3.1kinstalls5Wiki Lintwiki-lint is an Obsidian vault lint skill that runs structured quality checks across markdown pages in a personal or team wiki. It greps wikilink patterns, extracts link targets, and flags broken references when no matching file exists, offering rename updates, page creation, or link removal fixes. Frontmatter validation requires title, category, tags, sources, created, and updated on every page, with a soft warning when summary is missing or exceeds two hundred characters for cheap index-only retrieval. Stale content detection compares page updated timestamps against source file modification times. Contradiction checks read related tagged or cross-referenced pages for conflicting claims. Index consistency verifies index.md inventory matches files on disk and that listed summaries still match bodies. Provenance drift scoring counts extracted, inferred, and ambiguous markers, flagging speculation-heavy pages, unsourced synthesis, and hub pages with high inferred fractions. Fragmented tag cluster checks round out the lint pass. Developers use it to keep llm-ingested wikis trustworthy before publishing or sharing vault exports.3.1kinstalls6Wiki StatusWiki Status is an Obsidian wiki skill that reports ingest state, source-to-vault delta, and ranked next actions without modifying vault content except in insights mode. It resolves config for vault path, sources directory, and Claude or Codex history, then reads .manifest.json as the ingest ledger with canonical absolute source keys. It scans documents, Claude jsonl conversations, memory files, and Codex rollout transcripts, then classifies each source as new, modified, touched, unchanged, or deleted. The report includes page counts, visibility tallies from frontmatter tags, staged writes pending, and delta tables. Token footprint per tier uses a four chars per token heuristic with warnings when the full wiki exceeds WIKI_TOKEN_WARN_THRESHOLD. What to Do Next ranks up to six items from staged writes, _raw/ files, stale core pages, orphans, synthesis opportunities, and lint status. Insights mode builds a wikilink graph surfacing anchor hubs, bridge pages, tag cluster cohesion, surprising links, tier suggestions, and graph deltas in _insights.md.3kinstalls7Wiki Updatewiki-update is a cross-project agent skill for distilling the current working project into an Obsidian vault through the obsidian-wiki configuration protocol. It resolves OBSIDIAN_VAULT_PATH and related settings from project dotenv or home config, reads manifest.json and index.md for prior sync state, scans README docs source layout dependency manifests and meaningful git history, then computes a delta from last_commit_synced using merge-base validation with full rescan fallback after rebase or force-push. Distillation captures architecture decisions, reusable patterns, dependency wiring, trade-offs, and lessons not obvious from code while skipping boilerplate lockfiles and one-off bug fixes. New knowledge lands under projects project-name with overview concepts skills and references while global concepts skills entities and synthesis absorb reusable findings. Every page needs YAML frontmatter with provenance extracted inferred ambiguous fractions, lifecycle draft, wikilink cross-linking, manifest index log and hot.md updates, and optional QMD index refresh when QMD_WIKI_COLLECTION is set. Developers reach for it when users say update wiki, sync to wiki, or save project knowledge.3kinstalls8Cross Linkercross-linker automatically weaves an Obsidian wiki knowledge graph by discovering unlinked mentions of page names, titles, aliases, and entities, then inserting wikilinks or Related sections. It follows llm-wiki retrieval primitives: build a page registry from frontmatter greps, read full pages only for plausible link targets, and respect OBSIDIAN_LINK_FORMAT for wikilink versus markdown links. Scoring awards points for exact name matches, shared tags, same-project proximity, cross-category connections, and peripheral-to-hub reach, labeling candidates EXTRACTED, INFERRED, or AMBIGUOUS before applying links. Applied links infer relationship types such as uses, extends, contradicts, or related_to into frontmatter relationships blocks written in wikilink syntax regardless of body link format. A misc-page affinity pass flags promotion candidates when project connection scores reach three or more. The skill updates log.md and hot.md after writes, skips _archives content, and documents optional QMD update and embed refresh when QMD_WIKI_COLLECTION is configured after vault modifications.2.9kinstalls9Wiki Rebuildwiki-rebuild performs destructive Obsidian wiki operations with mandatory archiving first and explicit user confirmation before clearing live content. It resolves vault config via the llm-wiki protocol, reads .manifest.json for current state, and supports three modes: archive-only snapshot, archive plus full rebuild, or restore from a prior archive under _archives/. Archive directories copy category folders, index.md, log.md, projects, and manifest snapshots with archive-meta.json metadata. Rebuild mode archives with reason rebuild, clears live category directories while preserving _archives, .obsidian, and .env, resets index and log templates, and directs the user to run ingest skills rather than auto-ingesting. Restore lists available archives, archives current state as pre-restore, copies chosen archive content back, and suggests wiki-lint afterward. QMD refresh rules differ per mode, with rebuild and restore requiring update and embed when QMD_WIKI_COLLECTION is configured.2.9kinstalls10Wiki Exportwiki-export serializes an Obsidian wiki knowledge graph into external tool formats under wiki-export/ at the vault root. It globs markdown pages excluding archives and special files, builds node metadata from frontmatter id, title, category, tags, and summary, then extracts wikilink edges from page bodies plus typed relationships from frontmatter blocks. Optional project filters keep pages whose id starts with projects/name or whose tags include the project name. Visibility filtered mode excludes pages tagged visibility/internal or visibility/pii when users request public exports. Community ids cluster pages by dominant frontmatter tags for visualization coloring. Outputs include NetworkX-style graph.json, graph.graphml for Gephi and Cytoscape, cypher.txt MERGE statements for Neo4j with typed relationship labels, graph.html for browser visualization, and an optional OKF markdown bundle. The skill warns and stops when fewer than five pages exist and documents post-export QMD considerations separately from the graph files.2.9kinstalls11Wiki History Ingestwiki-history-ingest is a thin unified router for conversation and session history sources in the obsidian-wiki skill family. It does not replace wiki-ingest for documents; instead it dispatches explicit subcommands such as claude, copilot, codex, hermes, openclaw, or pi to their specialized history-ingest skills. Routing rules map filesystem paths like ~/.claude, ~/.copilot session-store.db, ~/.codex rollout artifacts, ~/.hermes memories, ~/.openclaw MEMORY.md files, and ~/.pi/agent/sessions JSONL to the correct destination skill. An auto subcommand infers the target from context using the same path rules when the user does not name an agent explicitly. When the target is ambiguous the router asks one short clarification question listing the supported agents. After routing it executes the destination workflow exactly without duplicating ingest logic, leaving manifest, index, and log updates to the specialized skill. UX convention separates wiki-ingest for document content sources from wiki-history-ingest for agent history sources.2.9kinstalls12Tag Taxonomytag-taxonomy is an Obsidian wiki skill that enforces a controlled tag vocabulary before creating or updating pages. It resolves vault config via the llm-wiki protocol, reads the canonical _meta/taxonomy.md list, and applies rules such as max five tags per page, lowercase hyphenated tags, broad-over-narrow preference, and alias normalization. Tag audit mode scans all markdown pages, builds frequency tables, and reports unknown tags, alias usage, over-tagged pages, and untagged pages while leaving visibility/ reserved tags untouched. Reserved visibility/public, visibility/internal, and visibility/pii tags are managed separately and do not count toward the five-tag limit. Normalization workflows map aliases to canonical forms and can extend the taxonomy when users request new approved tags. Developers reach for tag-taxonomy when tag sprawl breaks search, dashboards fail, or agents write inconsistent frontmatter across hundreds of notes in a living knowledge base.2.8kinstalls13Claude History IngestClaude History Ingest extracts knowledge from past Claude Code conversations into an Obsidian wiki vault with append-mode delta processing guided by .manifest.json and index.md state. Config resolution walks CWD .env, ~/.obsidian-wiki/config for OBSIDIAN_VAULT_PATH and CLAUDE_HISTORY_PATH defaulting to ~/.claude, with WIKI_SKIP_PROJECTS excluding matching project directories uniformly across scan and manifest steps. Pre-extraction via extract-jsonl.py strips tool_use, thinking, and progress noise for 50 to 200 times smaller files under ~/.claude/extracted, enabling deeper session coverage per token budget. Sampling reads memory/*.md first when present, then new conversations not in manifest, otherwise three most recent sessions per project by mtime, always reporting sampled versus skipped counts. Steps compile vault pages, update manifest ingested_at timestamps, and honor canonical absolute paths when comparing manifest keys. The skill triggers on process my Claude history, add conversations to wiki, or .claude folder mining requests and can route via wiki-history-ingest claude.2.6kinstalls14Skill Creatorskill-creator is an agent skill from ar9av/obsidian-wiki implemented as the Post-hoc Analyzer Agent that runs after a blind comparator picks winner A or B. The workflow unblinds results by reading winner_skill_path, loser_skill_path, winner_transcript_path, and loser_transcript_path, then explains why the winner succeeded and how to improve the loser with actionable edits. Developers reach for skill-creator when operating a skills library and needing evidence-backed refinement instead of guessing prompt tweaks. The analyzer focuses on transcript diffs, skill structure gaps, and repeatable patterns that made one execution outperform another. Use it after structured skill shootouts in Claude Code or Cursor when you already have comparison outputs and want a written improvement plan targeting the losing SKILL.md, not when bootstrapping a skill from scratch without A/B data.2.5kinstalls15Graph ColorizeThe graph-colorize skill rewrites an Obsidian vault .obsidian/graph.json colorGroups array to tint graph nodes by tag, folder, visibility, combined modes, or user-supplied custom mappings. It resolves OBSIDIAN_VAULT_PATH via the llm-wiki config protocol, verifies .obsidian exists, and warns that Obsidian overwrites graph.json on close so users should close the vault or reload immediately after edits. Modes include by-tag top ten tags excluding visibility system tags, by-category fixed folder palette for concepts through journal, by-visibility red orange green for pii internal public, combined visibility then tags, or custom hex mappings converted to packed rgb integers. A ten-color colorblind-friendly palette cycles when groups exceed ten entries. Merge step backs up graph.json to graph.json.backup-timestamp, replaces only colorGroups, and preserves zoom physics filters and display settings. Reports summarize mode, group count, palette, and backup path, appending GRAPH_COLORIZE lines to vault log.md. Edge cases fall back from empty tags to by-category, restore backups on undo, or clear colorGroups when requested.2.4kinstalls16Wiki CaptureThe wiki-capture skill preserves conversation substance as permanent Obsidian wiki notes, not chat transcripts. Full mode classifies content into synthesis, concept, source, decision, or session types, rewrites declarative knowledge with provenance markers, writes frontmatter-rich pages, links at least two existing wiki pages, and updates index.md, log.md, and hot.md. Quick mode via --quick drops topic-clustered findings to _raw/ in under sixty seconds with no manifest or index writes, gated by KEEP versus SKIP rules that favor manual invocation. Config resolves OBSIDIAN_VAULT_PATH and link format via the llm-wiki protocol. Full mode body templates vary by type with Context, Finding, Reasoning, and Related sections. Quick mode clusters findings by topic, assigns base_confidence from discussion through test confirmation, and defers promotion to /wiki-ingest. Optional QMD refresh runs update and embed after vault writes without rolling back on failure. Teams invoke it when users say save this, capture this, or file this conversation into lasting knowledge.2.4kinstalls17Wiki ResearchThe wiki-research skill runs a three-round research loop on a user topic, then files results into an Obsidian vault as permanent knowledge. Setup resolves OBSIDIAN_VAULT_PATH and OBSIDIAN_LINK_FORMAT, reads index.md and optional hot.md and research-config.md, and confirms ambiguous topics with the user. Round one decomposes the topic into three to five angles with WebSearch and WebFetch extraction capturing claims, concepts, entities, and contradictions. Round two fills gaps with up to five targeted searches. Round three resolves contradictions or halts when depth suffices after at most three rounds. Filing creates sources, concepts, and entities pages with provenance frontmatter plus synthesis/Research Topic.md as the master overview. Cross-linking requires concepts to cite sources while avoiding duplicates in index.md. Updates append research entries to .manifest.json, index.md, log.md, hot.md, and optional qmd update when QMD_WIKI_COLLECTION is set.2.4kinstalls18Wiki DashboardThe wiki-dashboard skill builds queryable Obsidian vault views using Obsidian Bases native YAML .base files or Dataview plugin SQL-like blocks. Config resolves OBSIDIAN_VAULT_PATH via the llm-wiki protocol before reading index.md for existing categories. Bases files use expression-string filters under and/or/not, never typed filter objects, with groupBy inside view definitions not top level. Property naming differs by context: note.tags in properties keys but bare tags in order and groupBy. Dataview supports TABLE, GROUP BY with rows. prefix, stale-page date math via file.mtime, and multi-folder FROM unions. Output lands in _meta/ with slugs like concepts-index.base or dashboard.md, optionally embedded via ![[_meta/name.base]]. Common recipes cover content indexes, entity card galleries, concepts by domain, ingestion logs, and thirty-day stale content trackers. log.md receives WIKI_DASHBOARD entries; optional QMD refresh follows vault markdown writes. Prefer Bases on Obsidian 1.8+ unless GROUP BY or computed columns require Dataview.2.4kinstalls19Wiki SynthesizeThe wiki-synthesize skill scans an Obsidian vault for concept pairs that co-occur across pages but lack a dedicated synthesis note. Resolve config via llm-wiki protocol for OBSIDIAN_VAULT_PATH and link format, then read index.md, hot.md, and _meta/taxonomy.md. Build a co-occurrence matrix counting pages linking to both concepts in each pair, targeting top 20-30 candidates via grep backlink intersection. Filter pairs already covered by existing synthesis/ pages reading sources frontmatter. Score remaining pairs by co-occurrence count, cross-domain categories, shared tags, hub flags, and contradiction resolution potential; draft the top five into synthesis/ using the A x B title template with inferred and ambiguous provenance markers. Each page needs Connection, Co-occur, Cross-cutting Insight, Tensions, Strongest Objection with testable query, and Open Questions sections. Back-link from source concept pages, update index.md and log.md, refresh hot.md activity, and optionally run qmd update plus embed when QMD_WIKI_COLLECTION is set. List the next ten skipped candidates for future runs. Synthesis must add insight beyond summarizing sources; skip pairs with no real conceptual link.2.4kinstalls20Codex History IngestThe codex-history-ingest skill mines past Codex CLI sessions and distills durable knowledge into an Obsidian wiki vault. It resolves config via OBSIDIAN_VAULT_PATH and CODEX_HISTORY_PATH defaulting to ~/.codex, reads .manifest.json and index.md before processing. Append mode ingests only new or modified rollout and index files; full mode reprocesses everything after wiki-rebuild. It surveys session_index.jsonl for inventory, parses rollout JSONL events prioritizing user intent and assistant messages while filtering telemetry, secrets, and raw system prompts. Knowledge clusters by topic across sessions into projects, concepts, skills, entities, and synthesis paths with provenance markers extracted, inferred, or ambiguous. Manifest updates track ingested_at, source_type, pages created or updated, and project summaries. After vault writes it optionally runs qmd update and embed for search index refresh. Privacy rules require redaction and distillation rather than verbatim transcript dumps.2.3kinstalls21Hermes History IngestHermes History Ingest Conversation Memory Mining You are extracting knowledge from the user s Hermes agent history and distilling it into the Obsidian wiki Hermes stores both free form memories and structured session transcripts focus on durable knowledge not operational telemetry This skill can be invoked directly or via the wiki history ingest router wiki history ingest hermes 1 Resolve config follow the Config Resolution Protocol in llm wiki SKILL md walk up CWD for env obsidian wiki config prompt setup This gives OBSIDIAN_VAULT_PATH and HERMES_HISTORY_PATH defaults to hermes 2 Read manifest json at the vault root to check what has already been ingested 3 Read index md at the vault root to understand what the wiki already contains Check manifest json for each source file Only process The hermes history ingest agent skill provides documented workflows prerequisites triggers and safety guidance from its SKILL md source Agents load it when user requests match the description and follow step by step instructions without inventing capabilities It integrates with standard agent2.1kinstalls22Wiki AgentWiki Agent Targeted Cross Agent History Search Ingest You are doing a query driven targeted ingest from one specific AI agent s raw conversation history The user is typically working in a different agent right now and wants to pull in context from another agent s past sessions This is not bulk ingest You find sessions about a specific topic extract the relevant blobs distill them into the wiki and return a synthesized answer the user can act on immediately Parse the invocation to determine the target agent and optional query Command Target Example wiki claude query Claude Code history wiki claude how did I set up auth middleware wiki codex query Codex CLI history wiki codex rust ownership patterns wiki hermes query Hermes agent history wiki hermes memory architecture wiki openclaw query OpenClaw history wiki openclaw project planning approach wiki copilot query Copilot chat history wiki copilot test strategy for API routes wiki pi query Pi agent history wiki pi how did I refactor the auth module2kinstalls23Daily UpdateYou run a lightweight maintenance pass over the wiki check source freshness refresh the index update hot md and write the state file that the terminal notification reads 1 Resolve config follow the Config Resolution Protocol in llm wiki SKILL md walk up CWD for env obsidian wiki config prompt setup This gives OBSIDIAN_VAULT_PATH and OBSIDIAN_WIKI_REPO 2 Derive vault scoped state dir all runtime state is scoped to the resolved vault not global bash VAULT_ID echo OBSIDIAN_VAULT_PATH md5sum 2 dev null cut c1 8 md5 q OBSIDIAN_VAULT_PATH cut c1 8 STATE_DIR HOME obsidian wiki state VAULT_ID mkdir p STATE_DIR 3 Read OBSIDIAN_VAULT_PATH manifest json Run Mode default triggered by cron or daily update Compare each source in manifest json against its file s modification time Classify as Fresh mtime ingested_at Stale mtime ingested_at new content exists not yet ingested Missing source file no longer exists Read OBSIDIAN_VAULT_PATH index md If any pages in the vault are missing from the index or vice versa update the index Use find OBSIDIAN_VAULT_PATH name2kinstalls24Memory BridgeThe memory-bridge skill helps developers browse and compare Obsidian wiki knowledge filtered by which AI tool originally produced it using manifest.json provenance and page sources frontmatter. Config resolution follows the llm-wiki SKILL.md walk-up .env and ~/.obsidian-wiki/config protocol for OBSIDIAN_VAULT_PATH. Modes include browse by tool, search topic within tool, diff unique pages and blind spots across tools, diff two specific tools, and full origin matrix map of every page per tool. Source map building maps claude_conversation, codex_rollout, hermes_memory, openclaw sessions, copilot transcripts, and ingest documents to tool buckets with pages_created and pages_updated sets. Diff mode is the killer feature surfacing knowledge gaps between tools the user may not know exist. Recognized tools: claude, codex, hermes, openclaw, copilot, pi, manual, ingest. Use for memory-bridge, cross-tool memory, compare AI tool memories, or what does hermes know that claude doesn't.2kinstalls25Openclaw History IngestThe openclaw-history-ingest skill Ingest OpenClaw agent history into the Obsidian wiki Use this skill when the user wants to mine their past OpenClaw sessions for knowledge import their openclaw folder extract insights from previous OpenClaw conversations or says things like process my OpenClaw history add my OpenClaw sessions to the wiki ingest openclaw or what have I worked on in OpenClaw Also triggers when OpenClaw stores both a structured long-term MEMORY md and per-session JSONL transcripts focus on durable knowledge not operational telemetry This skill can be invoked directly or via the router Resolve config follow the Config Resolution Protocol in walk up CWD for prompt setup Read at the vault root to check what has already been ingested 3 Read at the vault root to understand what the wiki already contains Ingest Modes Append Mode default Check for each source file Only process Files not in the manifest new session logs updated MEMORY md or daily notes Files whose modification time is newer than in the manifest Use this1.9kinstalls26Impl Validator> The impl-validator skill documents workflows and patterns from the repository SKILL.md. --- name: impl-validator description: > Validate whether an implementation matches its stated goal. Use this skill when a skill or agent wants a second opinion on its own output, when the user says "check this implementation", "validate what you did", "is this correct?", "review the output", or "did you do this right?". Also spawned automatically as a subagent by other skills (memory-bridge, daily-update) to self-check their outputs before presenting to the user. Returns a structured pass/warn/fail verdict with specific actionable issues. --- # Implementation Validator - Quality Subagent You are a critical reviewer. Another skill or agent has just done work and wants you to check it. Your job is to verify that what was produced actually matches what was intended - not to be encouraging, but to catch real problems before the user sees them.1.9kinstalls27Wiki SwitchThe wiki-switch skill manages multiple Obsidian wiki vault profiles stored as config files at ~/.obsidian-wiki/config.<name> with the active vault determined by a symlink at ~/.obsidian-wiki/config. Dispatch routes /wiki-switch name to switch, list to enumerate profiles, show to print a config with secrets redacted, and new to scaffold from the active template. Switch verifies the named config exists, runs ln -sf to repoint the symlink, reads OBSIDIAN_VAULT_PATH from the activated file, and confirms the vault path. List resolves the current symlink, reads optional comment headers as descriptions, and marks the active profile. New copies the active config, asks for vault-specific field updates while preserving global sections, updates the top comment, and does not auto-switch. API_KEY and SECRET lines are redacted in show output.1.8kinstalls28Wiki DigestThe wiki-digest skill from obsidian-wiki produces human-readable newsletter-style summaries of wiki knowledge activity, distinct from wiki-status which reports ingestion deltas. It resolves config via llm-wiki SKILL.md protocol (.env walk-up, ~/.obsidian-wiki/config, or setup prompt) for OBSIDIAN_VAULT_PATH and link format. Period parsing supports daily (24h), weekly (7d default), monthly (30d), ISO since dates, or explicit day counts. Steps read log.md (last 200 lines), hot.md session context, and optional _insights.md anchor pages; glob vault markdown skipping index, log, hot, AGENTS, _meta, _archives, _raw, and journal/digest pages. Classify pages as new (created in period) or updated. Output summarizes themes, connections, and review-worthy pages as a knowledge newsletter. Use for weekly summary, what did I learn, knowledge report, or /wiki-digest requests.1.8kinstalls29Copilot History Ingestcopilot-history-ingest is an agent skill from ar9av/obsidian-wiki that apply copilot-history-ingest agent skill workflows from documented skill.md guidance. # Copilot History Ingest — Conversation Mining You are extracting knowledge from the user's past GitHub Copilot CLI conversations and distilling it into the Obsidian wiki. Conversations are rich but messy — your job is to find the signal and compile it. This skill can be invoked directly or via the `wiki-history-ingest` router (`/wiki-history-ing Developers invoke copilot-history-ingest during build/integrations work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.5kinstalls30Wiki DedupThe wiki-dedup skill is designed for scan the Obsidian wiki for page-level identity collisions — different pages covering the same concept under different names (e.g. "RSC" vs "React Server Components") — and. Wiki Dedup — Identity Resolution and Page-Level Deduplication You are finding and merging wiki pages that cover the same concept under different names. This is a write-heavy, potentially destructive skill — page merges cannot be automatically undone. Invoke when the user asks about wiki dedup or related SKILL.md workflows.1.4kinstalls31Wiki Context PackThe wiki-context-pack skill Wiki Context Pack Bounded Token Retrieval You are producing a focused token-bounded context pack from the wiki Unlike wiki-query which answers a question this skill packages the most relevant wiki knowledge into a single markdown block that a downstream agent skill or user can consume directly Resolve config follow the Config Resolution Protocol in llm-wiki SKILL md walk up CWD for env obsidian-wiki config prompt setup This gives OBSIDIAN_VAULT_PATH and any QMD variables Read OBSIDIAN_VAULT_PATH hot md if it exists gives instant context on recent activity Read OBSIDIAN_VAULT_PATH index md the full page inventory Invocation Forms wiki-context-pack transformer attention mechanism budget 16000 wiki-context-pack my-project architecture decisions budget 8000 wiki-context-pack recent budget 4000 recent activity pack from hot md wiki-context-pack authentication patterns default budget 8000 tokens Parse the user's invocation to extract topic the query string required unless recent budget N token budget in tokens default 8000 max 100000 recent pack the most recently updated ingested pages instead of a topic query Algorithm Step 1 Relevance Pass cheap1.2kinstalls32Wiki Stage CommitThe wiki-stage-commit skill automates git staging and commit steps for Obsidian wiki repositories used as publishable knowledge bases. It helps agents inspect vault diffs, group related note changes, write clear commit messages, and commit without disturbing unrelated local edits. The workflow fits wikis that sync through git remotes where markdown notes, assets, and index pages change in batches before launch or distribution updates. Agents verify only intended paths are staged, avoid secrets or machine-specific config, and can follow repository conventions for conventional commit prefixes. Use when publishing Obsidian wiki updates that need disciplined git history rather than ad hoc saves.1.2kinstalls33Pi History Ingestpi-history-ingest is a session-mining skill for developers who use the Pi coding agent and maintain an Obsidian wiki. Pi sessions are stored as JSONL files under ~/.pi/agent/sessions; this skill extracts insights, decisions, and implementation notes from those logs and distills them into linked Obsidian pages. Developers invoke pi-history-ingest when they say process my Pi history, ingest ~/.pi, or ask what they have worked on in Pi. The workflow turns ephemeral agent conversations into durable, cross-linked documentation inside an existing wiki structure. It does not run Pi or replace session storage—it transforms historical Pi agent output into searchable knowledge for later reference, onboarding, and pattern reuse across projects.848installs34Wiki Importwiki-import is an agent skill in the ar9av/obsidian-wiki framework that imports knowledge from wiki-export into an Obsidian vault. It accepts two source types: graph.json produces page stubs with frontmatter, typed relationships, and Related wikilinks, while an OKF v0.1 bundle reconstructs full page bodies for lossless round-trips. The workflow runs six steps—source detection, conflict mode selection, map building, page reconstruction, vault metadata updates, and summary reporting—with merge, skip, and overwrite modes defaulting to merge. Developers reach for wiki-import when migrating vaults, loading graph.json exports, importing OKF bundles, or pairing with wiki-export output after exporting from another machine.683installs35Vault Skill FactoryThis skill generates a portable, self-contained Agent Skill from mature, curated Obsidian wiki pages, producing a SKILL.md plus a references/ folder in a review directory. A developer uses it to distill a cluster of verified notes into a reusable domain-expert skill for humans to inspect and optionally install. It is the inverse of wiki-capture and enforces strict guardrails against auto-installing or writing into global skill directories.414installs