By GillySpace27
A plain-markdown research-task vault: one file per project, Obsidian-Tasks emoji syntax, a daily journal, six-bucket proposal/project status, and a two-tier memory of people/projects/terms. Mirrors the capabilities of the Anthropic productivity plugin without flattening to a single TASKS.md. Bring your own people and projects — /start interviews you and seeds the vault.
Quick-capture a one-line task. Drops it into the most likely project file based on content; falls back to `inbox.md` if ambiguous. Ten-second rule — minimal questions, surgical edit.
Regenerate the interactive `dashboard.html` from the current vault. Runs `scripts/build_dashboard.py`, which parses every project `.md` file (skipping README/CLAUDE/projects/journaling/dashboard scaffolding) and renders a filterable board.
Append to today's daily note. Default appends a `## Work log` bullet; `--summary "..."` updates the headline; `--decision "..."` adds under Decisions & notes; `--followup "..."` adds under Follow-ups.
Initialize or open a research-task vault — on first run, create the folder, scaffold the core files, and interview you to seed people/projects/memory. On later runs, verify files and open today's note.
Six-bucket proposal/project status report — Untouched → Proposal Idea Scoped → Drafting Currently → Submitted and Pending → Awarded → Complete. Show every category even if empty.
Daily journaling rules for the research vault — one note per calendar day under `daily_notes/YYYY-MM-DD.md`, with stable headings (Summary, Work log, Tasks touched, Decisions & notes, Follow-ups). Use whenever attempting or completing a task in the vault, and always as part of the "update the vault" convention. Append surgically; never rewrite or reorder.
Two-tier memory for the research vault. The vault `CLAUDE.md` is the hot cache (~30 most-used people, terms, projects). The `memory/` directory holds full storage — `glossary.md`, `people/<name>.md`, `projects/<name>.md`, `context/`. Use this skill whenever decoding shorthand, recording a new person/project/term, or deciding whether something belongs in the hot cache vs. deep storage.
Canonical roster of people the user works with — full names, affiliations, role on which projects, and the short handles the user writes (e.g.
Project model for the research vault — one `.md` file per project, listed in `projects.md`. Use whenever a task or new line of work needs routing to (or creating as) a project file, or when the user refers to a project by a short codename or acronym. The user's actual project list and codenames live in `projects.md` and the vault `CLAUDE.md` hot cache — read those for the live roster.
Six-bucket status grouping for the user's proposals and projects. Use whenever the user asks "what is the status of my proposals/projects", "where do things stand", or any variant. Sort and group into Untouched → Proposal Idea Scoped → Drafting Currently → Submitted and Pending → Awarded → Complete (or otherwise no further work). Show every category even if empty. The live proposal list lives in `proposal-solicitations.md`.
External network access
Connects to servers outside your machine
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub gillyspace27/research-vault-plugin --plugin research-vaultBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Claude Code plugin that turns a folder of plain-markdown files into a research-task vault: one file per project, GitHub-style checkbox tasks with Obsidian Tasks-plugin emoji metadata, a daily journal, a six-bucket proposal/project status view, and a two-tier memory of your people, projects, and shorthand.
It mirrors the capabilities of Anthropic's official productivity plugin
(/start, /update, /update --comprehensive, memory tiering, MCP-backed
external sync, an HTML dashboard) but keeps the per-project markdown shape, the
emoji task syntax, the daily journal, and the proposal status groupings —
none of which fit the single-TASKS.md model.
Bring your own everything. The plugin ships conventions, not anyone's
data. /start creates the vault and interviews you to seed your own people,
projects, and memory.
/plugin marketplace add GillySpace27/research-vault-plugin
/plugin install research-vault@research-vault-local
(Or point the marketplace at a local clone:
/plugin marketplace add /path/to/research-vault-plugin.)
Then, in any Claude Code session:
/start
On first run, /start asks where the vault should live (default
~/research-vault/), creates the folder, scaffolds the core files, and runs a
short interview to seed your roster, projects, and glossary. Nothing leaves
your machine.
New to it? GETTING-STARTED.md is the first-hour walkthrough.
/dashboard (the HTML board). The vault itself
works without it.dashboard.md a live
board and renders the emoji metadata./update --comprehensive.By default the vault lives at ~/research-vault/. To use a fixed location
elsewhere, either pass it (/start /path/to/vault) or set an environment
variable so /start and /dashboard find it automatically:
export RESEARCH_VAULT_DIR=~/path/to/vault
| Command | What it does |
|---|---|
/start | First run: create the vault + interview to seed it. Later: verify files, open today's note. |
/update | "Update the vault": reconcile every project file relevant to the recent conversation, advance/close tasks, append to today's daily note. |
/update --comprehensive | Plus deep scan: Slack/Gmail/Calendar/Notion/project-tracker for missed tasks, status changes, new people. |
/status | Six-bucket proposal/project status report (Untouched → Complete). |
/journal | Append a ## Work log line (or fuller entry) to today's daily note. |
/triage | Inbox triage: walk inbox.md, decide a route, move surgically. |
/capture <text> | Drop a one-line task into the most likely project file (or inbox.md if ambiguous). 10-second rule. |
/dashboard | Regenerate the interactive dashboard.html from the vault. |
Loaded automatically when you're working in the vault — they encode the conventions so future sessions don't need to be re-briefed.
| Skill | Role |
|---|---|
vault-tasks | Task syntax (checkboxes + emoji metadata), capture rules. |
vault-projects | Project manifest, when to make a new file, operating principles. |
vault-people | Roster decoder for handles like #firstname-lastname. Your roster lives in the vault CLAUDE.md hot cache. |
vault-journal | Daily journal location, structure, when/how to update. |
vault-update | The "update the vault" multi-step reconciliation flow. |
vault-status | Six-bucket status grouping for proposals/projects. |
vault-memory | Two-tier memory (vault CLAUDE.md hot cache + memory/ deep storage). |
<vault>/
├── CLAUDE.md ← hot-cache memory (your people/projects/prefs) + plugin pointer
├── README.md ← human-readable vault overview
├── dashboard.md ← Obsidian Tasks queries
├── dashboard.html ← (optional) interactive board view, via /dashboard
├── projects.md ← project manifest
├── journaling.md ← daily journal guidelines
├── inbox.md ← capture catch-all
├── personal.md ← non-work
├── proposal-solicitations.md ← funding calls (read by /status)
├── proposal-ideas.md ← idea-first proposal capture
├── <project>.md ← one per project
├── daily_notes/YYYY-MM-DD.md
├── memory/ ← two-tier memory (hot cache lives in CLAUDE.md)
│ ├── glossary.md
│ ├── people/<name>.md
│ ├── projects/<name>.md
│ └── context/
└── _templates/ ← project.md, daily-note.md, glossary.md, dashboard.md, …
The _examples/ folder in the plugin shows the intended shape of a filled-in
roster, manifest, and hot cache (using fictional names) — read them for
orientation; /start writes your real versions.
Harness-native ECC plugin for engineering teams - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows