From ed055b71aaf95a6da97819599850ba3319d05743 Mon Sep 17 00:00:00 2001 From: cc Date: Fri, 10 Apr 2026 15:39:42 -0400 Subject: [PATCH 1/2] feat: rewrite managing-obsidian-notes skill for public use (v2.0) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - Full onboarding: app install → CLI enable → vault create → PARA scaffold - Uses Obsidian CLI instead of direct filesystem access - Vault bootstrap with PARA-inspired structure (Projects/Areas/Resources/Archive) - Per-project folder scaffolding with templates - Always pops open notes in Obsidian after create/update - Starter templates for projects, papers, meetings - No hardcoded paths — works with any user's vault - Added to skills catalog and all pipeline stages Co-Authored-By: Claude Opus 4.6 (1M context) --- skills/managing-obsidian-notes/CHANGELOG.md | 31 ++ skills/managing-obsidian-notes/SKILL.md | 459 ++++++++++++++++++++ skills/skills-catalog-v2.json | 43 +- skills/stage-skill-map.json | 17 +- 4 files changed, 542 insertions(+), 8 deletions(-) create mode 100644 skills/managing-obsidian-notes/CHANGELOG.md create mode 100644 skills/managing-obsidian-notes/SKILL.md diff --git a/skills/managing-obsidian-notes/CHANGELOG.md b/skills/managing-obsidian-notes/CHANGELOG.md new file mode 100644 index 00000000..fb5fccfe --- /dev/null +++ b/skills/managing-obsidian-notes/CHANGELOG.md @@ -0,0 +1,31 @@ +# Changelog + +## 2026-04-10 (v2.0) +- **Rewritten for Obsidian CLI**: All operations now use `Obsidian` CLI instead of direct filesystem access +- **Setup warmup**: Detects missing CLI, guides user through installation +- **Vault discovery**: Auto-discovers vaults via `Obsidian vaults verbose` instead of hardcoding paths +- **Open in Obsidian**: Always offers to pop open notes in a new tab after create/update +- **iCloud sync guard**: `sleep 1` before opening notes after edits to prevent sync race +- **No personal paths**: Removed all hardcoded vault paths — works with any user's vault +- **Public-ready**: Suitable for bundling in dr-claw plugin + +## 2025-01-26 (v1.3) +- **Fixed vault path**: Corrected to `iCloud~md~obsidian/Documents/kobo-note/` +- **Added vault structure**: Documented all main folders (Career, Coding, Engineering, Research, Journal, etc.) +- **Enhanced CREATE NOTE routing**: Added folder mappings for all content types + +## 2025-12-13 (v1.2) +- **Chinese Headline Summary (总结)**: Content-rich subpages MUST start with `## 总结` in Chinese +- Added to quality checklist + +## 2025-12-13 (v1.1) +- **Index Page Philosophy**: Index pages = MINIMAL, Subpages = STRUCTURED +- Index pages should be simple navigation hubs, easy to manually edit +- All detailed content goes in subpages + +## 2025-12-13 (v1.0) +- Initial release +- CRUD operations for Obsidian vault notes +- Folder-with-index pattern support +- Obsidian syntax reference (wikilinks, callouts, tags, frontmatter) +- Vault location: kobo-note in iCloud diff --git a/skills/managing-obsidian-notes/SKILL.md b/skills/managing-obsidian-notes/SKILL.md new file mode 100644 index 00000000..9c23df38 --- /dev/null +++ b/skills/managing-obsidian-notes/SKILL.md @@ -0,0 +1,459 @@ +--- +name: managing-obsidian-notes +description: >- + CRUD operations for Obsidian vault notes using the Obsidian CLI. Use when asked to "create a note", + "add a folder", "update notes", "organize notes", "save to obsidian", "open in obsidian", or manage + any Obsidian/markdown note structure. Handles folder-with-index-file pattern, wikilinks, tags, + frontmatter, and Obsidian-specific syntax. Requires Obsidian v1.8+ with CLI enabled. +--- + +# Managing Obsidian Notes — Execution Rules + +## Goal & Scope +Perform Create, Read, Update, Delete operations on Obsidian vault notes using the Obsidian CLI, +following consistent patterns and Obsidian best practices. + +## Audience & Tone +Users with Obsidian vaults who want consistent note organization. Technical, concise. + +## Setup & Onboarding + +Follow these steps in order. Skip any step where the check already passes. + +### Step 1: Check if Obsidian App is Installed + +```bash +ls /Applications/Obsidian.app 2>/dev/null && echo "APP_OK" || echo "APP_MISSING" +``` + +If `APP_MISSING`, guide the user: + +> **Obsidian is not installed.** Let's set it up: +> +> **macOS:** +> ```bash +> brew install --cask obsidian +> ``` +> Or download from [obsidian.md/download](https://obsidian.md/download) +> +> **Linux:** +> ```bash +> # Snap +> sudo snap install obsidian --classic +> # Or AppImage from obsidian.md/download +> ``` +> +> **Windows:** +> Download installer from [obsidian.md/download](https://obsidian.md/download) +> +> After installing, **open Obsidian once** to complete initial setup, then continue. + +### Step 2: Check CLI Availability + +```bash +Obsidian version 2>/dev/null && echo "CLI_OK" || echo "CLI_MISSING" +``` + +If `CLI_MISSING` but app is installed: + +> **Obsidian CLI needs to be enabled:** +> +> 1. Open Obsidian +> 2. Go to **Settings → General** +> 3. Scroll down and enable **"Enable CLI"** +> 4. **Restart Obsidian** (quit and reopen) +> 5. Run `Obsidian version` to verify +> +> The CLI requires Obsidian v1.8+. If you don't see the option, update Obsidian first. + +**Important:** Obsidian must be running for the CLI to work. The CLI talks to the app process. + +### Step 3: Vault Discovery + +Check if the user has any vaults: + +```bash +Obsidian vaults verbose +``` + +**If vaults exist:** Ask the user which one to use. Store the vault name for the session. + +**If no vaults exist:** Guide vault creation: + +> **No vaults found.** Let's create one: +> +> 1. Open Obsidian +> 2. Click **"Create new vault"** +> 3. Name it (e.g. "research", "notes", or your project name) +> 4. Choose a location (default is fine, or pick a cloud-synced folder) +> 5. Click **Create** +> +> Then run `Obsidian vaults verbose` again and tell me the vault name. + +**Store the vault name for the session** — don't ask again after the first command. + +### Step 4: Vault Structure Bootstrap + +After discovering the vault, check if it has structure: + +```bash +Obsidian vault= folders +``` + +If the vault is **empty or flat** (no folders, or only `.obsidian/` and `.trash/`), offer to +bootstrap a research-friendly structure. + +> Your vault is empty. I can set up a starter structure designed for research and project work. +> It follows a lightweight PARA-inspired layout: +> +> ``` +> / +> ├── Home.md ← Dashboard with quick links +> ├── Projects/ ← Active research projects (one folder per project) +> ├── Areas/ ← Ongoing responsibilities (teaching, advising, admin) +> ├── Resources/ ← Reference material, learning notes, how-tos +> │ ├── Papers/ ← Paper reading notes +> │ └── Learning/ ← Tutorials, courses, concepts +> ├── Archive/ ← Completed/inactive projects and notes +> ├── Inbox/ ← Quick capture — sort later +> ├── Templates/ ← Reusable note templates +> └── Attachments/ ← Images, PDFs, media +> ``` +> +> Want me to create this? + +If the user agrees, create the scaffold: + +```bash +# Home — vault dashboard +Obsidian vault= create name="Home" content="# Home\n\nDashboard for your research vault.\n\n## Navigation\n- [[Projects/Projects|Projects]] — active research\n- [[Areas/Areas|Areas]] — ongoing responsibilities\n- [[Resources/Resources|Resources]] — reference material\n- [[Inbox/Inbox|Inbox]] — quick capture\n- [[Archive/Archive|Archive]] — completed work\n\n## Recent Projects\n\n## Quick Notes\n" silent + +# Projects hub +Obsidian vault= create name="Projects" folder="Projects" content="# Projects\n\nOne folder per active research project.\n\n## Active\n\n## On Hold\n\n---\nSee [[Archive/Archive|Archive]] for completed projects.\n" silent + +# Areas hub +Obsidian vault= create name="Areas" folder="Areas" content="# Areas\n\nOngoing responsibilities with no end date.\n\n## Research\n\n## Teaching\n\n## Service\n" silent + +# Resources hub +Obsidian vault= create name="Resources" folder="Resources" content="# Resources\n\nReference material and learning notes.\n\n- [[Resources/Papers/Papers|Papers]] — reading notes\n- [[Resources/Learning/Learning|Learning]] — tutorials & concepts\n" silent + +# Papers subfolder +Obsidian vault= create name="Papers" folder="Resources/Papers" content="# Papers\n\nReading notes and paper summaries. One note per paper.\n\n## By Topic\n\n## Recent\n" silent + +# Learning subfolder +Obsidian vault= create name="Learning" folder="Resources/Learning" content="# Learning\n\nConcepts, tutorials, courses, and study notes.\n" silent + +# Archive +Obsidian vault= create name="Archive" folder="Archive" content="# Archive\n\nCompleted or paused projects and old notes.\nMove items here instead of deleting.\n" silent + +# Inbox +Obsidian vault= create name="Inbox" folder="Inbox" content="# Inbox\n\nQuick capture — review and sort weekly.\n" silent + +# Templates folder (empty index) +Obsidian vault= create name="Templates" folder="Templates" content="# Templates\n\nReusable note templates.\n" silent + +# Attachments folder marker +Obsidian vault= create name="Attachments" folder="Attachments" content="# Attachments\n\nImages, PDFs, and media files.\nSet this as the attachment folder in Settings → Files & Links.\n" silent +``` + +Then create starter templates: + +```bash +# Project template +Obsidian vault= create name="Project Template" folder="Templates" content="# {{title}}\n\n> [!info] Created {{date}}\n\n## Goal\n\n## Pages\n- [[system-design]]\n- [[progress]]\n- [[key-links]]\n- [[experiment-log]]\n\n## Related Papers\n\n## Notes\n" silent + +# Paper note template +Obsidian vault= create name="Paper Template" folder="Templates" content="---\ntitle: \"{{title}}\"\nauthors: \"\"\nyear: \nvenue: \"\"\nurl: \"\"\ntags: [paper]\n---\n\n# {{title}}\n\n## Key Contributions\n\n## Method\n\n## Results\n\n## Relevance\n" silent + +# Meeting note template +Obsidian vault= create name="Meeting Template" folder="Templates" content="# Meeting: {{title}}\n\n**Date:** {{date}}\n**Attendees:**\n\n## Agenda\n\n## Notes\n\n## Action Items\n- [ ] \n" silent +``` + +Finally, open Home: +```bash +sleep 1 && Obsidian vault= open newtab file="Home" +``` + +Tell the user: +> Vault is ready! I set up a PARA-style structure: +> - **Projects/** — create a subfolder for each research project +> - **Resources/Papers/** — one note per paper you read +> - **Inbox/** — quick capture, sort weekly +> - **Templates/** — I added templates for projects, papers, and meetings +> +> Tip: In Obsidian Settings → Templates, set the template folder to `Templates/`. + +### Per-Project Folder Structure + +When the user starts a new research project, create a dedicated project folder: + +```bash +# Project index (minimal — just links) +Obsidian vault= create name="" folder="Projects/" content="# \n\n> [!info] Created YYYY-MM-DD\n\n## Pages\n- [[system-design]]\n- [[progress]]\n- [[key-links]]\n- [[experiment-log]]\n\n## Related Papers\n" silent + +# System design doc +Obsidian vault= create name="system-design" folder="Projects/" content="# System Design\n\n## Architecture\n\n## Key Decisions\n" silent + +# Progress tracker +Obsidian vault= create name="progress" folder="Projects/" content="# Progress\n\n## Timeline\n\n### YYYY-MM-DD\n- Project started\n" silent + +# Key links (repo, server, docs) +Obsidian vault= create name="key-links" folder="Projects/" content="# Key Links\n\n| Resource | Link |\n|----------|------|\n| GitHub | |\n| Server | |\n| Paper | |\n| Data | |\n" silent + +# Experiment log +Obsidian vault= create name="experiment-log" folder="Projects/" content="# Experiment Log\n\nAppend entries as experiments run.\n" silent +``` + +Set properties on the project index: +```bash +Obsidian vault= property:set name="created" value="YYYY-MM-DD" file="" +Obsidian vault= property:set name="status" value="active" file="" +Obsidian vault= property:set name="tags" value="project" file="" +``` + +Then add it to the Projects hub and open: +```bash +Obsidian vault= append file="Projects" content="\n- [[Projects/|]]" +sleep 1 && Obsidian vault= open newtab file="" +``` + +## CLI Command Reference + +All operations use the `Obsidian` binary (capital O on macOS). Commands follow: +```bash +Obsidian vault= [params...] +``` + +### CRUD Operations + +**Create:** +```bash +Obsidian vault= create name="Note Title" content="# Note Title\n\nContent here" silent +# With folder: +Obsidian vault= create name="Note Title" folder="Projects/Active" content="..." silent +# Overwrite existing: +Obsidian vault= create name="Note Title" content="..." silent overwrite +``` + +**Read:** +```bash +# By wikilink name: +Obsidian vault= read file="Note Title" +# By exact path: +Obsidian vault= read path="Projects/Active/plan.md" +``` + +**Append:** +```bash +Obsidian vault= append file="Note Title" content="\n## New Section\nContent here" +``` + +**Search:** +```bash +Obsidian vault= search query="search terms" limit=10 +# JSON output for programmatic use: +Obsidian vault= search query="search terms" format=json +``` + +**List files:** +```bash +Obsidian vault= files folder="Projects" +``` + +**List folders:** +```bash +Obsidian vault= folders folder="Projects" +``` + +### Opening Notes in Obsidian + +**ALWAYS use `newtab`** so the note opens in a new tab, not replacing the current one: + +```bash +Obsidian vault= open newtab file="Note Title" +Obsidian vault= open newtab path="Projects/Active/plan.md" +``` + +**After creating or editing a note, ALWAYS offer to open it:** +``` +Created "Note Title". Open in Obsidian? (I'll pop it up in a new tab) +``` + +If the user says yes: +```bash +sleep 1 && Obsidian vault= open newtab file="Note Title" +``` + +> **Why `sleep 1`?** If the vault is on iCloud/Dropbox, the file watcher can race with sync. +> Without the delay, Obsidian may show the old cached version. + +### Properties (Frontmatter) + +```bash +# Get all properties: +Obsidian vault= properties file="Note Title" format=json + +# Set a property: +Obsidian vault= property:set name="status" value="active" file="Note Title" +``` + +### Tags & Backlinks + +```bash +# List all tags with counts: +Obsidian vault= tags format=json counts + +# Tags for a specific note: +Obsidian vault= tags file="Note Title" format=json + +# Backlinks (who links to this note): +Obsidian vault= backlinks file="Note Title" format=json +``` + +### Tasks + +```bash +# List incomplete tasks: +Obsidian vault= tasks todo format=json + +# List completed tasks: +Obsidian vault= tasks done format=json + +# Toggle a task: +Obsidian vault= task file="Note Title" line=15 toggle +``` + +### Vault Info + +```bash +Obsidian vault= vault +``` + +## Core Patterns + +### Folder-with-Index Pattern +When creating a topic that needs multiple sub-notes: +``` +Topic Name/ +├── Topic Name.md # Index file (same name as folder) - MINIMAL +├── Subtopic 1.md # Detailed content +├── Subtopic 2.md # Detailed content +└── Resources/ +``` + +### Index Page Philosophy +> **Index pages = MINIMAL** | **Subpages = STRUCTURED** + +**Index pages:** Simple list of wikilinks to subpages — just navigation. +**Subpages:** Detailed and well-organized — tables, callouts, checklists. + +### Single Note Pattern +For standalone topics, just create a single `.md` file in the appropriate location. + +## Required Structure / Algorithm + +### CREATE NOTE +1. Check CLI is available (run `Obsidian version`) +2. If first time, discover vault (`Obsidian vaults verbose`) +3. If vault is empty/unstructured, offer to bootstrap (see Vault Structure Bootstrap above) +4. If user mentions a "project", use Per-Project Folder Structure to create the full scaffold +5. Determine folder based on content type (ask user if unclear) +4. Decide pattern: single note vs folder-with-index +5. Create with CLI: + ```bash + Obsidian vault= create name="Title" folder="Target/Folder" content="..." silent + ``` +6. Set frontmatter properties if needed: + ```bash + Obsidian vault= property:set name="created" value="2026-04-10" file="Title" + Obsidian vault= property:set name="tags" value="tag1, tag2" file="Title" + ``` +7. **Offer to open in Obsidian** — always ask, then: + ```bash + sleep 1 && Obsidian vault= open newtab file="Title" + ``` + +### READ/FIND NOTE +1. Search by content: + ```bash + Obsidian vault= search query="keyword" format=json limit=10 + ``` +2. Read specific note: + ```bash + Obsidian vault= read file="Note Title" + ``` +3. Browse folder: + ```bash + Obsidian vault= files folder="Projects/Active" + ``` + +### UPDATE NOTE +1. Read existing content first: + ```bash + Obsidian vault= read file="Note Title" + ``` +2. For appending: + ```bash + Obsidian vault= append file="Note Title" content="\n## New Section\nContent" + ``` +3. For full rewrite: + ```bash + Obsidian vault= create name="Note Title" content="..." silent overwrite + ``` +4. Update properties: + ```bash + Obsidian vault= property:set name="modified" value="2026-04-10" file="Note Title" + ``` + +### DELETE/ARCHIVE NOTE +1. Prefer archiving over deletion — move to an archive folder +2. For deletion, confirm with user first +3. Use filesystem operations for moves (CLI doesn't have a move command): + ```bash + mv "/Notes/old.md" "/.archive/old.md" + ``` + +## Obsidian Syntax Quick Reference + +| Syntax | Example | +|--------|---------| +| Internal link | `[[Note Name]]` | +| Aliased link | `[[Note Name\|Display Text]]` | +| Heading link | `[[Note Name#Heading]]` | +| Embed note | `![[Note Name]]` | +| Tag | `#tag/subtag` | +| Callout | `> [!note] Title` | +| Task | `- [ ] Todo` / `- [x] Done` | + +## Quality Checklist +- [ ] CLI is available (`Obsidian version` returns version) +- [ ] Vault is specified (not hardcoded) +- [ ] Frontmatter is valid YAML (if used) +- [ ] Internal links use `[[wikilink]]` syntax +- [ ] Folder pattern used for multi-part topics +- [ ] Index file has same name as parent folder +- [ ] Offered to open note in Obsidian after create/update + +## Failure Modes & Recovery + +**CLI not installed** +→ Show setup instructions (see Setup & Warmup section) + +**Obsidian not running** +→ Tell user: "Obsidian needs to be running for the CLI to work. Please open Obsidian first." + +**Vault not found** +→ Run `Obsidian vaults verbose` and ask user to pick from the list + +**Note already exists** +→ Offer to update existing or create with different name. Use `overwrite` flag if user confirms. + +**iCloud/Dropbox sync race** +→ Always `sleep 1` before opening a note after create/edit + +## Limits +- Only manage markdown/Obsidian files +- Don't modify `.obsidian/` config folder +- Don't delete without explicit confirmation +- Don't create files outside the vault +- CLI requires Obsidian v1.8+ to be running diff --git a/skills/skills-catalog-v2.json b/skills/skills-catalog-v2.json index 07da0969..d4121ae9 100644 --- a/skills/skills-catalog-v2.json +++ b/skills/skills-catalog-v2.json @@ -1,7 +1,7 @@ { "generatedAt": "2026-03-24T17:51:17.721Z", "schema": "skills-taxonomy-v2", - "totalSkills": 170, + "totalSkills": 171, "skills": [ { "name": "academic-researcher", @@ -7456,6 +7456,45 @@ "ds-write", "ds-review" ] + }, + { + "name": "managing-obsidian-notes", + "primaryIntent": "research", + "intents": [ + "research", + "writing" + ], + "capabilities": [ + "infrastructure-ops" + ], + "domains": [ + "general" + ], + "keywords": [ + "obsidian", + "notes", + "vault", + "knowledge-management", + "markdown", + "wikilinks", + "create-note", + "read-note", + "search-notes", + "obsidian-cli", + "backlinks", + "tags", + "frontmatter", + "project-notes" + ], + "source": "imported", + "status": "verified", + "summary": "CRUD operations for Obsidian vault notes using the Obsidian CLI. Handles vault bootstrap for new users, per-project folder scaffolding, folder-with-index pattern, wikilinks, tags, frontmatter, and pops open notes in Obsidian after create/update. Requires Obsidian v1.8+ with CLI enabled.", + "legacy": { + "dirPath": "managing-obsidian-notes", + "skillFile": "skills/managing-obsidian-notes/SKILL.md", + "topLevelGroup": "Standalone", + "subGroup": null + } } ] -} \ No newline at end of file +} diff --git a/skills/stage-skill-map.json b/skills/stage-skill-map.json index e4786f5a..c487bff3 100644 --- a/skills/stage-skill-map.json +++ b/skills/stage-skill-map.json @@ -8,7 +8,8 @@ "biorxiv-database", "aris-idea-discovery", "aris-research-lit", - "ds-scout" + "ds-scout", + "managing-obsidian-notes" ], "byTaskType": { "analysis": [ @@ -43,7 +44,8 @@ "academic-researcher", "aris-idea-discovery", "aris-idea-creator", - "ds-idea" + "ds-idea", + "managing-obsidian-notes" ], "byTaskType": { "analysis": [ @@ -81,7 +83,8 @@ "aris-run-experiment", "autoresearch", "ds-experiment", - "ds-baseline" + "ds-baseline", + "managing-obsidian-notes" ], "byTaskType": { "implementation": [ @@ -126,7 +129,8 @@ "aris-rebuttal", "ds-write", "ds-review", - "ds-rebuttal" + "ds-rebuttal", + "managing-obsidian-notes" ], "byTaskType": { "writing": [ @@ -164,7 +168,8 @@ "base": [ "making-academic-presentations", "aris-paper-poster", - "aris-paper-slides" + "aris-paper-slides", + "managing-obsidian-notes" ], "byTaskType": { "scripting": [ @@ -183,4 +188,4 @@ ] } } -} \ No newline at end of file +} From 6c18d14d195113f467bda4929644898160523af3 Mon Sep 17 00:00:00 2001 From: cc Date: Fri, 10 Apr 2026 15:42:25 -0400 Subject: [PATCH 2/2] docs: add managing-obsidian-notes to skill taxonomy, index, and README - Add Knowledge Management section to skills/README.md taxonomy tree - Add entry to .agents/skills/skills-index.md - Update README.md plugin section to mention MCP servers Co-Authored-By: Claude Opus 4.6 (1M context) --- .agents/skills/skills-index.md | 253 +++++++++++++++++++++++++++++++++ README.md | 3 +- skills/README.md | 3 + 3 files changed, 258 insertions(+), 1 deletion(-) create mode 100644 .agents/skills/skills-index.md diff --git a/.agents/skills/skills-index.md b/.agents/skills/skills-index.md new file mode 100644 index 00000000..7a5e42e3 --- /dev/null +++ b/.agents/skills/skills-index.md @@ -0,0 +1,253 @@ +# Skills Index + +> **Do NOT read all SKILL.md files at once.** Use this index to find the right skill, then read only that one. + +## Core Pipeline Skills + +| Skill | Path | Description | +|-------|------|-------------| +| dataset-discovery | `.agents/skills/dataset-discovery/SKILL.md` | Multi-source ML dataset discovery. Search HuggingFace Hub, OpenML, GitHub, and paper cross-references for datasets re... | +| init-analysis | `.agents/skills/init-analysis/SKILL.md` | This skill should be used when the user asks to "run initial analysis", "analyze single-cell data", "QC my data", "ru... | +| inno-code-survey | `.agents/skills/inno-code-survey/SKILL.md` | Acquires missing code repositories for the selected idea (Phase A) and conducts comprehensive code survey mapping aca... | +| inno-deep-research | `.agents/skills/inno-deep-research/SKILL.md` | Comprehensive research assistant that synthesizes information from multiple sources with citations. Use when: conduct... | +| inno-experiment-analysis | `.agents/skills/inno-experiment-analysis/SKILL.md` | This skill should be used when the user asks to "analyze experimental results", "generate results section", "statisti... | +| inno-experiment-dev | `.agents/skills/inno-experiment-dev/SKILL.md` | Creates implementation plan, writes project code with judge feedback loop, and submits final experiment run. Use afte... | +| inno-figure-gen | `.agents/skills/inno-figure-gen/SKILL.md` | Generate/edit images with Gemini image models (default: gemini-3.1-flash-image-preview). Use for image create/modify ... | +| inno-grant-proposal | `.agents/skills/inno-grant-proposal/SKILL.md` | Help professors and researchers write, revise, adapt, and polish grant proposals for US agencies (NSF, NIH, DOE, DARP... | +| inno-humanizer | `.agents/skills/inno-humanizer/SKILL.md` | Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and ... | +| inno-idea-eval | `.agents/skills/inno-idea-eval/SKILL.md` | Multi-persona idea evaluation with quality gate. Evaluates ideas across 5 InnoEval dimensions (Clarity, Novelty, Vali... | +| inno-idea-generation | `.agents/skills/inno-idea-generation/SKILL.md` | Facilitates structured brainstorming sessions, conducts comprehensive research, and generates creative solutions usin... | +| inno-paper-reviewer | `.agents/skills/inno-paper-reviewer/SKILL.md` | Structured manuscript/grant review with checklist-based evaluation. Use when writing formal peer reviews with specifi... | +| inno-paper-writing | `.agents/skills/inno-paper-writing/SKILL.md` | Creates formal academic research papers following IEEE/ACM formatting standards with proper structure, citations, and... | +| managing-obsidian-notes | `skills/managing-obsidian-notes/SKILL.md` | Obsidian vault CRUD via CLI. Vault bootstrap (PARA structure), per-project scaffolding, templates. Pops open notes in Obsidian after create/update. | +| inno-pipeline-planner | `.agents/skills/inno-pipeline-planner/SKILL.md` | Guides the user through an interactive conversation to define their research project, then generates research_brief.j... | +| inno-prepare-resources | `.agents/skills/inno-prepare-resources/SKILL.md` | Loads the evaluation instance, searches GitHub for related repositories, builds a dataset description, queries the Pr... | +| inno-rclone-to-overleaf | `.agents/skills/inno-rclone-to-overleaf/SKILL.md` | Access Overleaf projects via CLI. Use for reading/writing LaTeX files, syncing local .tex files to Overleaf, download... | +| inno-rebuttal | `.agents/skills/inno-rebuttal/SKILL.md` | Drafting and refining academic rebuttals for top-tier AI/CS conferences (NeurIPS, ICML, ICLR, CVPR, ECCV, AAAI, ARR, ... | +| inno-reference-audit | `.agents/skills/inno-reference-audit/SKILL.md` | This skill provides reference guidance for citation verification in academic writing. Use when the user asks about "c... | + +## Library Skills + +| Skill | Path | Description | +|-------|------|-------------| +| ml-paper-writing | `.agents/skills/library/20-ml-paper-writing/SKILL.md` | Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research... | +| academic-researcher | `.agents/skills/library/academic-researcher/SKILL.md` | Academic research assistant for literature reviews, paper analysis, and scholarly writing. Use when: reviewing academ... | +| huggingface-accelerate | `.agents/skills/library/accelerate/SKILL.md` | Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpee... | +| huggingface-accelerate | `.agents/skills/library/accelerate/SKILL.md` | Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpee... | +| adding-leetcode-templates | `.agents/skills/library/adding-leetcode-templates/SKILL.md` | Add algorithm/data structure template code to the leetcode-template Anki deck with proper HTML formatting. Use this s... | +| ai-reference-audit | `.agents/skills/library/ai-reference-audit/SKILL.md` | Run strong reference audits for AI research papers and repos. Use when asked to check LaTeX/BibTeX citations for miss... | +| audiocraft-audio-generation | `.agents/skills/library/audiocraft/SKILL.md` | PyTorch library for audio generation including text-to-music (MusicGen) and text-to-sound (AudioGen). Use when you ne... | +| audiocraft-audio-generation | `.agents/skills/library/audiocraft/SKILL.md` | PyTorch library for audio generation including text-to-music (MusicGen) and text-to-sound (AudioGen). Use when you ne... | +| autogpt-agents | `.agents/skills/library/autogpt/SKILL.md` | Autonomous AI agent platform for building and deploying continuous agents. Use when creating visual workflow agents, ... | +| autogpt-agents | `.agents/skills/library/autogpt/SKILL.md` | Autonomous AI agent platform for building and deploying continuous agents. Use when creating visual workflow agents, ... | +| awq-quantization | `.agents/skills/library/awq/SKILL.md` | Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when de... | +| awq-quantization | `.agents/skills/library/awq/SKILL.md` | Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when de... | +| axolotl | `.agents/skills/library/axolotl/SKILL.md` | Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimo... | +| axolotl | `.agents/skills/library/axolotl/SKILL.md` | Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimo... | +| batch-condensing-arxiv-papers | `.agents/skills/library/batch-condensing-arxiv-papers/SKILL.md` | Downloads and condenses multiple arXiv papers to their essential LaTeX content. Use when the user requests to "downlo... | +| evaluating-code-models | `.agents/skills/library/bigcode-evaluation-harness/SKILL.md` | Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when ... | +| evaluating-code-models | `.agents/skills/library/bigcode-evaluation-harness/SKILL.md` | Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when ... | +| biorxiv-database | `.agents/skills/library/biorxiv-database/SKILL.md` | Efficient database search tool for bioRxiv preprint server. Use this skill when searching for life sciences preprints... | +| quantizing-models-bitsandbytes | `.agents/skills/library/bitsandbytes/SKILL.md` | Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limit... | +| quantizing-models-bitsandbytes | `.agents/skills/library/bitsandbytes/SKILL.md` | Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limit... | +| blip-2-vision-language | `.agents/skills/library/blip-2/SKILL.md` | Vision-language pre-training framework bridging frozen image encoders and LLMs. Use when you need image captioning, v... | +| blip-2-vision-language | `.agents/skills/library/blip-2/SKILL.md` | Vision-language pre-training framework bridging frozen image encoders and LLMs. Use when you need image captioning, v... | +| brainstorming-research-ideas | `.agents/skills/library/brainstorming-research-ideas/SKILL.md` | Guides researchers through structured ideation frameworks to discover high-impact research directions. Use when explo... | +| chroma | `.agents/skills/library/chroma/SKILL.md` | Open-source embedding database for AI applications. Store embeddings and metadata, perform vector and full-text searc... | +| chroma | `.agents/skills/library/chroma/SKILL.md` | Open-source embedding database for AI applications. Store embeddings and metadata, perform vector and full-text searc... | +| clip | `.agents/skills/library/clip/SKILL.md` | OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross... | +| clip | `.agents/skills/library/clip/SKILL.md` | OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross... | +| constitutional-ai | `.agents/skills/library/constitutional-ai/SKILL.md` | Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with s... | +| constitutional-ai | `.agents/skills/library/constitutional-ai/SKILL.md` | Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with s... | +| converting-pptx-to-images | `.agents/skills/library/converting-pptx-to-images/SKILL.md` | Convert a PPTX file into a folder of per-slide PNG images using LibreOffice and pdftoppm. Use when the user asks to "... | +| creative-thinking-for-research | `.agents/skills/library/creative-thinking-for-research/SKILL.md` | Applies cognitive science frameworks for creative thinking to CS and AI research ideation. Use when seeking genuinely... | +| crewai-multi-agent | `.agents/skills/library/crewai/SKILL.md` | Multi-agent orchestration framework for autonomous AI collaboration. Use when building teams of specialized agents wo... | +| crewai-multi-agent | `.agents/skills/library/crewai/SKILL.md` | Multi-agent orchestration framework for autonomous AI collaboration. Use when building teams of specialized agents wo... | +| cross-machine-sync | `.agents/skills/library/cross-machine-sync/SKILL.md` | Sync and push files, text, and clipboard between MacBook Pro and Mac Mini over Tailscale SSH. Use when user asks to "... | +| deepspeed | `.agents/skills/library/deepspeed/SKILL.md` | Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/F... | +| deepspeed | `.agents/skills/library/deepspeed/SKILL.md` | Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/F... | +| dspy | `.agents/skills/library/dspy/SKILL.md` | Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and... | +| dspy | `.agents/skills/library/dspy/SKILL.md` | Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and... | +| excalidraw | `.agents/skills/library/excalidraw/SKILL.md` | Create and edit Excalidraw diagrams programmatically. Use when the user asks to create diagrams, flowcharts, architec... | +| experimental-rigor-audit | `.agents/skills/library/experimental-rigor-audit/SKILL.md` | Audit experiments and writeups for scientific rigor, reasoning transparency, and common pitfalls. Use when reviewing ... | +| faiss | `.agents/skills/library/faiss/SKILL.md` | Facebook's library for efficient similarity search and clustering of dense vectors. Supports billions of vectors, GPU... | +| faiss | `.agents/skills/library/faiss/SKILL.md` | Facebook's library for efficient similarity search and clustering of dense vectors. Supports billions of vectors, GPU... | +| optimizing-attention-flash | `.agents/skills/library/flash-attention/SKILL.md` | Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/... | +| optimizing-attention-flash | `.agents/skills/library/flash-attention/SKILL.md` | Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/... | +| gemini-deep-research | `.agents/skills/library/gemini-deep-research/SKILL.md` | Perform deep, multi-source research using Google Gemini's Deep Research Agent. Use this skill whenever the user asks ... | +| gguf-quantization | `.agents/skills/library/gguf/SKILL.md` | GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardwar... | +| gguf-quantization | `.agents/skills/library/gguf/SKILL.md` | GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardwar... | +| gptq | `.agents/skills/library/gptq/SKILL.md` | Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on c... | +| gptq | `.agents/skills/library/gptq/SKILL.md` | Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on c... | +| grpo-rl-training | `.agents/skills/library/grpo-rl-training/SKILL.md` | Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training | +| grpo-rl-training | `.agents/skills/library/grpo-rl-training/SKILL.md` | Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training | +| guidance | `.agents/skills/library/guidance/SKILL.md` | Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and... | +| guidance | `.agents/skills/library/guidance/SKILL.md` | Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and... | +| hqq-quantization | `.agents/skills/library/hqq/SKILL.md` | Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision with... | +| hqq-quantization | `.agents/skills/library/hqq/SKILL.md` | Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision with... | +| huggingface-tokenizers | `.agents/skills/library/huggingface-tokenizers/SKILL.md` | Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Suppor... | +| huggingface-tokenizers | `.agents/skills/library/huggingface-tokenizers/SKILL.md` | Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Suppor... | +| instructor | `.agents/skills/library/instructor/SKILL.md` | Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse co... | +| instructor | `.agents/skills/library/instructor/SKILL.md` | Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse co... | +| invoking-codex | `.agents/skills/library/invoking-codex/SKILL.md` | Invoke Codex CLI as a subagent to perform tasks and return results. Use when you want a second opinion, need Codex's ... | +| invoking-gemini | `.agents/skills/library/invoking-gemini/SKILL.md` | Invoke Gemini CLI as a subagent to perform tasks and return results. Use when you want a second opinion from Gemini, ... | +| json-canvas | `.agents/skills/library/json-canvas/SKILL.md` | Create and edit JSON Canvas files (.canvas) with nodes, edges, groups, and connections. Use when working with .canvas... | +| knowledge-distillation | `.agents/skills/library/knowledge-distillation/SKILL.md` | Compress large language models using knowledge distillation from teacher to student models. Use when deploying smalle... | +| knowledge-distillation | `.agents/skills/library/knowledge-distillation/SKILL.md` | Compress large language models using knowledge distillation from teacher to student models. Use when deploying smalle... | +| lambda-labs-gpu-cloud | `.agents/skills/library/lambda-labs/SKILL.md` | Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances w... | +| lambda-labs-gpu-cloud | `.agents/skills/library/lambda-labs/SKILL.md` | Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances w... | +| langchain | `.agents/skills/library/langchain/SKILL.md` | Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, An... | +| langchain | `.agents/skills/library/langchain/SKILL.md` | Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, An... | +| langsmith-observability | `.agents/skills/library/langsmith/SKILL.md` | LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating m... | +| langsmith-observability | `.agents/skills/library/langsmith/SKILL.md` | LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating m... | +| lehigh-aisp-gpu | `.agents/skills/library/lehigh-aisp-gpu/SKILL.md` | Connect to Lehigh AISP GPU server (aisp.cse.lehigh.edu) as yila22 with key ~/.ssh/id_rsa_unify, and quickly browse co... | +| implementing-llms-litgpt | `.agents/skills/library/litgpt/SKILL.md` | Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mi... | +| implementing-llms-litgpt | `.agents/skills/library/litgpt/SKILL.md` | Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mi... | +| llama-cpp | `.agents/skills/library/llama-cpp/SKILL.md` | Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M... | +| llama-cpp | `.agents/skills/library/llama-cpp/SKILL.md` | Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M... | +| llama-factory | `.agents/skills/library/llama-factory/SKILL.md` | Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimod... | +| llama-factory | `.agents/skills/library/llama-factory/SKILL.md` | Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimod... | +| llamaguard | `.agents/skills/library/llamaguard/SKILL.md` | Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual ... | +| llamaguard | `.agents/skills/library/llamaguard/SKILL.md` | Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual ... | +| llamaindex | `.agents/skills/library/llamaindex/SKILL.md` | Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing,... | +| llamaindex | `.agents/skills/library/llamaindex/SKILL.md` | Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing,... | +| llava | `.agents/skills/library/llava/SKILL.md` | Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP v... | +| llava | `.agents/skills/library/llava/SKILL.md` | Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP v... | +| evaluating-llms-harness | `.agents/skills/library/lm-evaluation-harness/SKILL.md` | Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking ... | +| evaluating-llms-harness | `.agents/skills/library/lm-evaluation-harness/SKILL.md` | Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking ... | +| long-context | `.agents/skills/library/long-context/SKILL.md` | Extend context windows of transformer models using RoPE, YaRN, ALiBi, and position interpolation techniques. Use when... | +| long-context | `.agents/skills/library/long-context/SKILL.md` | Extend context windows of transformer models using RoPE, YaRN, ALiBi, and position interpolation techniques. Use when... | +| mac-mini-ssh | `.agents/skills/library/mac-mini-ssh/SKILL.md` | Connect to Mac Mini via SSH over Tailscale. Use for running commands remotely, syncing files between MacBook Pro and ... | +| macbook-ssh | `.agents/skills/library/macbook-ssh/SKILL.md` | Connect to MacBook Pro via SSH over Tailscale. Use for running commands remotely, syncing files between Mac Mini and ... | +| making-academic-presentations | `.agents/skills/library/making-academic-presentations/SKILL.md` | Create academic presentation slide decks and optionally demo videos from research papers. Use when the user asks to "... | +| making-nsf-pose-insight-slides | `.agents/skills/library/making-nsf-pose-insight-slides/SKILL.md` | Create NSF-POSE insight slides as a PPTX. Use when the user asks for "NSF-POSE slides", "POSE insight PPTX", "insight... | +| making-skills | `.agents/skills/library/making-skills/SKILL.md` | Build new Claude Skills following best practices. Use this skill when the user asks to "create a skill", "make a new ... | +| mamba-architecture | `.agents/skills/library/mamba/SKILL.md` | State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV ca... | +| mamba-architecture | `.agents/skills/library/mamba/SKILL.md` | State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV ca... | +| managing-calendar-events | `.agents/skills/library/managing-calendar-events/SKILL.md` | Manage Google Calendar events using gcalcli. Use this skill when the user asks to "create a calendar event", "add eve... | +| managing-cli-wrapper-apis | `.agents/skills/library/managing-cli-wrapper-apis/SKILL.md` | Set up and reuse local OpenAI-compatible API servers that wrap Codex CLI or Claude Code. Use this when asked to copy ... | +| managing-github-repos | `.agents/skills/library/managing-github-repos/SKILL.md` | Manage GitHub repositories using the gh CLI. Use this skill when the user asks to "create a repo", "clone a repo", "l... | +| managing-obsidian-notes | `.agents/skills/library/managing-obsidian-notes/SKILL.md` | CRUD operations for Obsidian vault notes. Use when asked to "create a note", "add a folder", "update notes", "organiz... | +| managing-omnifocus-tasks | `.agents/skills/library/managing-omnifocus-tasks/SKILL.md` | Manage OmniFocus tasks using AppleScript/osascript. Use this skill when the user asks to "add task to OmniFocus", "cr... | +| managing-python-packages | `.agents/skills/library/managing-python-packages/SKILL.md` | Manage Python package installations using uv (ultra-fast Python package installer). Use this skill when the user asks... | +| managing-uf-gpu | `.agents/skills/library/managing-uf-gpu/SKILL.md` | | +| training-llms-megatron | `.agents/skills/library/megatron-core/SKILL.md` | Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Us... | +| training-llms-megatron | `.agents/skills/library/megatron-core/SKILL.md` | Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Us... | +| miles-rl-training | `.agents/skills/library/miles/SKILL.md` | Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training l... | +| miles-rl-training | `.agents/skills/library/miles/SKILL.md` | Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training l... | +| ml-paper-writing | `.agents/skills/library/ml-paper-writing/SKILL.md` | Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research... | +| mlflow | `.agents/skills/library/mlflow/SKILL.md` | Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments w... | +| mlflow | `.agents/skills/library/mlflow/SKILL.md` | Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments w... | +| modal-serverless-gpu | `.agents/skills/library/modal/SKILL.md` | Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure... | +| modal-serverless-gpu | `.agents/skills/library/modal/SKILL.md` | Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure... | +| model-merging | `.agents/skills/library/model-merging/SKILL.md` | Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating special... | +| model-merging | `.agents/skills/library/model-merging/SKILL.md` | Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating special... | +| model-pruning | `.agents/skills/library/model-pruning/SKILL.md` | Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing mode... | +| model-pruning | `.agents/skills/library/model-pruning/SKILL.md` | Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing mode... | +| moe-training | `.agents/skills/library/moe-training/SKILL.md` | Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limit... | +| moe-training | `.agents/skills/library/moe-training/SKILL.md` | Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limit... | +| nano-banana | `.agents/skills/library/nano-banana/SKILL.md` | Render images from text prompts via PaperBanana (multi-agent academic figure pipeline) or Gemini CLI's nanobanana ext... | +| nanobanana-prompt-architect | `.agents/skills/library/nanobanana-prompt-architect/SKILL.md` | Generate high-fidelity, math-accurate prompts for academic framework figures (ICML/NeurIPS style) and optionally rend... | +| nanogpt | `.agents/skills/library/nanogpt/SKILL.md` | Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learni... | +| nanogpt | `.agents/skills/library/nanogpt/SKILL.md` | Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learni... | +| nemo-curator | `.agents/skills/library/nemo-curator/SKILL.md` | GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× fa... | +| nemo-curator | `.agents/skills/library/nemo-curator/SKILL.md` | GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× fa... | +| nemo-evaluator-sdk | `.agents/skills/library/nemo-evaluator/SKILL.md` | Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend exe... | +| nemo-evaluator-sdk | `.agents/skills/library/nemo-evaluator/SKILL.md` | Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend exe... | +| nemo-guardrails | `.agents/skills/library/nemo-guardrails/SKILL.md` | NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-c... | +| nemo-guardrails | `.agents/skills/library/nemo-guardrails/SKILL.md` | NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-c... | +| nnsight-remote-interpretability | `.agents/skills/library/nnsight/SKILL.md` | Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote ... | +| nnsight-remote-interpretability | `.agents/skills/library/nnsight/SKILL.md` | Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote ... | +| obsidian-bases | `.agents/skills/library/obsidian-bases/SKILL.md` | Create and edit Obsidian Bases (.base files) with views, filters, formulas, and summaries. Use when working with .bas... | +| obsidian-markdown | `.agents/skills/library/obsidian-markdown/SKILL.md` | Create and edit Obsidian Flavored Markdown with wikilinks, embeds, callouts, properties, and other Obsidian-specific ... | +| openrlhf-training | `.agents/skills/library/openrlhf/SKILL.md` | High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B... | +| openrlhf-training | `.agents/skills/library/openrlhf/SKILL.md` | High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B... | +| outlines | `.agents/skills/library/outlines/SKILL.md` | Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local m... | +| outlines | `.agents/skills/library/outlines/SKILL.md` | Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local m... | +| paper-analyzer | `.agents/skills/library/paper-analyzer/SKILL.md` | Deep analysis of a single paper — generate structured notes with figures, evaluation, and knowledge graph updates | +| paper-finder | `.agents/skills/library/paper-finder/SKILL.md` | Search existing paper notes by title, author, keyword, or research domain | +| paper-image-extractor | `.agents/skills/library/paper-image-extractor/SKILL.md` | Extract figures from papers — prioritizes arXiv source package for high-quality images | +| peft-fine-tuning | `.agents/skills/library/peft/SKILL.md` | Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70... | +| peft-fine-tuning | `.agents/skills/library/peft/SKILL.md` | Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70... | +| phoenix-observability | `.agents/skills/library/phoenix/SKILL.md` | Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM application... | +| phoenix-observability | `.agents/skills/library/phoenix/SKILL.md` | Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM application... | +| pinecone | `.agents/skills/library/pinecone/SKILL.md` | Managed vector database for production AI applications. Fully managed, auto-scaling, with hybrid search (dense + spar... | +| pinecone | `.agents/skills/library/pinecone/SKILL.md` | Managed vector database for production AI applications. Fully managed, auto-scaling, with hybrid search (dense + spar... | +| pose-interview-logging | `.agents/skills/library/pose-interview-logging/SKILL.md` | Log POSE program interviews into the Innovation Within discovery platform using Playwright automation. Creates interv... | +| pose-schedule | `.agents/skills/library/pose-schedule/SKILL.md` | Add POSE program tasks to OmniFocus and Google Calendar, and create session calendar invitations with Zoom links and ... | +| processing-forum-posts | `.agents/skills/library/processing-forum-posts/SKILL.md` | Process and clean forum posts from 一亩三分地 (1point3acres) for interview preparation. Use this skill when the user asks ... | +| processing-gemini-talks | `.agents/skills/library/processing-gemini-talks/SKILL.md` | Split Gemini conversation exports into individual knowledge notes and place them in the Obsidian vault. Use when the ... | +| prompt-guard | `.agents/skills/library/prompt-guard/SKILL.md` | Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+... | +| prompt-guard | `.agents/skills/library/prompt-guard/SKILL.md` | Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+... | +| pytorch-fsdp2 | `.agents/skills/library/pytorch-fsdp2/SKILL.md` | Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and... | +| pytorch-fsdp2 | `.agents/skills/library/pytorch-fsdp2/SKILL.md` | Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and... | +| pytorch-lightning | `.agents/skills/library/pytorch-lightning/SKILL.md` | High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks syste... | +| pytorch-lightning | `.agents/skills/library/pytorch-lightning/SKILL.md` | High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks syste... | +| pyvene-interventions | `.agents/skills/library/pyvene/SKILL.md` | Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention frame... | +| pyvene-interventions | `.agents/skills/library/pyvene/SKILL.md` | Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention frame... | +| qdrant-vector-search | `.agents/skills/library/qdrant/SKILL.md` | High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG system... | +| qdrant-vector-search | `.agents/skills/library/qdrant/SKILL.md` | High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG system... | +| ray-data | `.agents/skills/library/ray-data/SKILL.md` | Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Inte... | +| ray-data | `.agents/skills/library/ray-data/SKILL.md` | Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Inte... | +| ray-train | `.agents/skills/library/ray-train/SKILL.md` | Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nod... | +| ray-train | `.agents/skills/library/ray-train/SKILL.md` | Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nod... | +| research-landscape-assessment | `.agents/skills/library/research-landscape-assessment/SKILL.md` | Assess the research landscape for a given topic: quantify hotness/saturation, identify sub-directions, find key paper... | +| research-news | `.agents/skills/library/research-news/SKILL.md` | Daily paper recommendation workflow — search arXiv and Semantic Scholar, score and recommend papers | +| rwkv-architecture | `.agents/skills/library/rwkv/SKILL.md` | RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), in... | +| rwkv-architecture | `.agents/skills/library/rwkv/SKILL.md` | RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), in... | +| sparse-autoencoder-training | `.agents/skills/library/saelens/SKILL.md` | Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network act... | +| sparse-autoencoder-training | `.agents/skills/library/saelens/SKILL.md` | Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network act... | +| scientific-writing | `.agents/skills/library/scientific-writing/SKILL.md` | Core skill for the deep research and writing tool. Write scientific manuscripts in full paragraphs (never bullet poin... | +| searching-agent-sessions | `.agents/skills/library/searching-agent-sessions/SKILL.md` | Search and browse past coding agent sessions (Claude Code, Codex, Gemini) via CASS CLI. Use when the user asks to "fi... | +| searching-ai-papers | `.agents/skills/library/searching-ai-papers/SKILL.md` | Search general-domain AI/ML papers across arXiv, Semantic Scholar, OpenAlex, OpenReview, ICLR accepted papers (OpenRe... | +| searching-local-papers | `.agents/skills/library/searching-local-papers/SKILL.md` | Search the local corpus of 115K+ AI/ML research papers (79% with abstracts, 84% for 2025+ papers) using a precomputed... | +| searching-rednote-posts | `.agents/skills/library/searching-rednote-posts/SKILL.md` | Search Xiaohongshu (小红书/RedNote) posts using Playwright MCP browser automation. Use when the user asks to "search Red... | +| segment-anything-model | `.agents/skills/library/segment-anything/SKILL.md` | Foundation model for image segmentation with zero-shot transfer. Use when you need to segment any object in images us... | +| segment-anything-model | `.agents/skills/library/segment-anything/SKILL.md` | Foundation model for image segmentation with zero-shot transfer. Use when you need to segment any object in images us... | +| sending-lehigh-emails | `.agents/skills/library/sending-lehigh-emails/SKILL.md` | Send emails using Lehigh University Gmail account (yila22@lehigh.edu) via Gmail MCP. Use this skill when the user ask... | +| sentence-transformers | `.agents/skills/library/sentence-transformers/SKILL.md` | Framework for state-of-the-art sentence, text, and image embeddings. Provides 5000+ pre-trained models for semantic s... | +| sentence-transformers | `.agents/skills/library/sentence-transformers/SKILL.md` | Framework for state-of-the-art sentence, text, and image embeddings. Provides 5000+ pre-trained models for semantic s... | +| sentencepiece | `.agents/skills/library/sentencepiece/SKILL.md` | Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences... | +| sentencepiece | `.agents/skills/library/sentencepiece/SKILL.md` | Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences... | +| session-migrate | `.agents/skills/library/session-migrate/SKILL.md` | Migrate Claude Code sessions between MacBook Pro and Mac Mini over Tailscale SSH. Remaps user paths so sessions creat... | +| sglang | `.agents/skills/library/sglang/SKILL.md` | Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, const... | +| sglang | `.agents/skills/library/sglang/SKILL.md` | Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, const... | +| simpo-training | `.agents/skills/library/simpo/SKILL.md` | Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 poi... | +| simpo-training | `.agents/skills/library/simpo/SKILL.md` | Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 poi... | +| skypilot-multi-cloud-orchestration | `.agents/skills/library/skypilot/SKILL.md` | Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or bat... | +| skypilot-multi-cloud-orchestration | `.agents/skills/library/skypilot/SKILL.md` | Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or bat... | +| slime-rl-training | `.agents/skills/library/slime/SKILL.md` | Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM model... | +| slime-rl-training | `.agents/skills/library/slime/SKILL.md` | Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM model... | +| speculative-decoding | `.agents/skills/library/speculative-decoding/SKILL.md` | Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use wh... | +| speculative-decoding | `.agents/skills/library/speculative-decoding/SKILL.md` | Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use wh... | +| stable-diffusion-image-generation | `.agents/skills/library/stable-diffusion/SKILL.md` | State-of-the-art text-to-image generation with Stable Diffusion models via HuggingFace Diffusers. Use when generating... | +| stable-diffusion-image-generation | `.agents/skills/library/stable-diffusion/SKILL.md` | State-of-the-art text-to-image generation with Stable Diffusion models via HuggingFace Diffusers. Use when generating... | +| syncing-overleaf-papers | `.agents/skills/library/syncing-overleaf-papers/SKILL.md` | Two-way file management between local and remote Dropbox/Overleaf via rclone. Use when the user asks to "sync to over... | +| syncing-skill-hub | `.agents/skills/library/syncing-skill-hub/SKILL.md` | Check skill sync status and optionally back up to git. Use when the user asks to "sync skills", "save skills", "push ... | +| tensorboard | `.agents/skills/library/tensorboard/SKILL.md` | Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile pe... | +| tensorboard | `.agents/skills/library/tensorboard/SKILL.md` | Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile pe... | +| tensorrt-llm | `.agents/skills/library/tensorrt-llm/SKILL.md` | Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment... | +| tensorrt-llm | `.agents/skills/library/tensorrt-llm/SKILL.md` | Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment... | +| toggling-mac-displays | `.agents/skills/library/toggling-mac-displays/SKILL.md` | Toggle physical monitors on/off for Mac Mini to enable virtual displays for remote desktop access. Use when asked to ... | +| torchforge-rl-training | `.agents/skills/library/torchforge/SKILL.md` | Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Us... | +| torchforge-rl-training | `.agents/skills/library/torchforge/SKILL.md` | Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Us... | +| distributed-llm-pretraining-torchtitan | `.agents/skills/library/torchtitan/SKILL.md` | Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use whe... | +| distributed-llm-pretraining-torchtitan | `.agents/skills/library/torchtitan/SKILL.md` | Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use whe... | +| transformer-lens-interpretability | `.agents/skills/library/transformer-lens/SKILL.md` | Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transform... | +| transformer-lens-interpretability | `.agents/skills/library/transformer-lens/SKILL.md` | Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transform... | +| trending-research-scanner | `.agents/skills/library/trending-research-scanner/SKILL.md` | Iterative field radar — scan trending research papers across Semantic Scholar, OpenAlex, HF Daily Papers, and arXiv w... | +| fine-tuning-with-trl | `.agents/skills/library/trl-fine-tuning/SKILL.md` | Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/... | +| fine-tuning-with-trl | `.agents/skills/library/trl-fine-tuning/SKILL.md` | Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/... | +| unsloth | `.agents/skills/library/unsloth/SKILL.md` | Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization | +| unsloth | `.agents/skills/library/unsloth/SKILL.md` | Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization | +| updating-leetcode-anki-cards | `.agents/skills/library/updating-leetcode-anki-cards/SKILL.md` | Transform LeetCode Anki flashcards from code-implementation-heavy format to semantic-reasoning-focused format followi... | +| verl-rl-training | `.agents/skills/library/verl/SKILL.md` | Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing... | +| verl-rl-training | `.agents/skills/library/verl/SKILL.md` | Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing... | +| serving-llms-vllm | `.agents/skills/library/vllm/SKILL.md` | Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production L... | +| serving-llms-vllm | `.agents/skills/library/vllm/SKILL.md` | Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production L... | +| weights-and-biases | `.agents/skills/library/weights-and-biases/SKILL.md` | Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, a... | +| weights-and-biases | `.agents/skills/library/weights-and-biases/SKILL.md` | Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, a... | +| whisper | `.agents/skills/library/whisper/SKILL.md` | OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and ... | +| whisper | `.agents/skills/library/whisper/SKILL.md` | OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and ... | +| writing-experiment-reports | `.agents/skills/library/writing-experiment-reports/SKILL.md` | Generate comprehensive LaTeX experiment reports for machine learning ablation studies. Use when user requests "create... | +| writing-standup-notes | `.agents/skills/library/writing-standup-notes/SKILL.md` | Create structured daily standup notes following the Yesterday/Today/Blockers format. Use this skill when the user ask... | diff --git a/README.md b/README.md index 1518afdb..1708c1ab 100644 --- a/README.md +++ b/README.md @@ -824,7 +824,8 @@ The plugin is installed and will be available in future sessions. Run `/drclaw:s ### What's Included -- **60+ bundled skills** — a curated subset of Dr. Claw's skill library covering literature survey, idea generation, experiment development, paper writing, and more +- **60+ bundled skills** — a curated subset of Dr. Claw's skill library covering literature survey, idea generation, experiment development, paper writing, knowledge management, and more +- **2 MCP servers** — `drclaw-papers` for multi-source paper search (arXiv, Semantic Scholar, OpenAlex, OpenReview, HF Daily) and `drclaw-notes` for Obsidian vault operations via CLI - **3 project templates** — Method/Model, Dataset/Benchmark, and Position Paper, each with a pre-configured task pipeline - **Auto-detection** — the plugin detects existing pipeline projects on session start and shows your current progress - **Same data format** — projects created with the plugin use the same `research_brief.json` and `tasks.json` schemas as the full Dr. Claw workspace, so you can switch between them diff --git a/skills/README.md b/skills/README.md index 29ae97ae..d9f71f69 100644 --- a/skills/README.md +++ b/skills/README.md @@ -55,6 +55,9 @@ skills/ ├─ Promotion │ └─ making-academic-presentations Slides, narration, TTS audio, and demo-video generation │ +├─ Knowledge Management +│ └─ managing-obsidian-notes Obsidian vault CRUD via CLI, vault bootstrap, project scaffolding +│ └─ Domain-Specific └─ bioinformatics-init-analysis CyTOF / scRNA-seq / flow cytometry pipeline ```