# Lit Status

> Lit Status

- Skill: `the-fish-salmon/lit-status` (Agent Skill)
- Install (CLI): `npx skillmds@latest add the-fish-salmon/lit-status`
- Raw SKILL.md: https://api.skillmd.com/api/skills/the-fish-salmon/lit-status/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: The-Fish-Salmon (https://skillmd.com/u/the-fish-salmon)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/the-fish-salmon/lit-status

---


# Lit Status

A read-only skill that surfaces what's in `30_Literature/` and how it relates to the user's active projects.

## What it answers

- **How many papers are in the library**, broken down by `status` (unread / skimmed / read / cited / dismissed) and by `year`.
- **What tags / topics** are present, and how concentrated.
- **What's been read recently** (last 30 / 90 days, by `added` and by `status` change time).
- **What papers are cited in active project notes but missing from the library** (so the user can capture them).
- **What papers in the library don't link to any project** (orphaned reading).
- **Reading queue**: unread papers, ranked by how often their citekey is referenced in active project notes.

## Workflow

1. Use the `obsidian` MCP to enumerate notes in `30_Literature/`. Read frontmatter for each (citekey, status, tags, projects, added, year, doi, arxiv).
2. Use `obsidian` MCP search to find citekey references across `10_Projects/`, `60_MOCs/`, `00_Inbox/`. Build a citation count map.
3. Compose the requested report.

## Modes

The skill supports several report shapes; the slash command or invocation passes a mode hint:

- **`summary`** (default) -- one-paragraph overview + counts.
- **`unread`** -- list of unread papers, ranked by project-reference count.
- **`tags`** -- tag cloud + papers per tag.
- **`gaps`** -- papers cited in projects but not captured.
- **`orphans`** -- papers in library not linked to any project or MOC.
- **`citation-map {citekey}`** -- for one paper, show which project notes reference it.
- **`next-action`** -- "what should I read next?" Returns a single, prioritized next action in <=5 lines: the most-cited unread paper in the active project (or, if no active project, the most-recently-added unread paper), plus the exact slash command the user can run to act on it (typically `/capture-paper` or `/copilot synthesize {citekey1} {citekey2}`). Used by `research-copilot` for its Suggest move.

## Constraints

- **Read-only.** Never writes to the vault. If the user wants to update a paper's status, point them to the file path; don't edit it for them.
- If the `obsidian` MCP is offline, fall back to filesystem `Glob` + `Grep` against the vault path from `OBSIDIAN_VAULT_PATH`.
- Limit each report section to <=30 entries; if there are more, show top 30 and the total count.

## Output style

Markdown tables for lists; one paragraph for summaries. Always include the count at the top of each table so the user knows whether they're seeing all or a subset.

## Example output

```
## Library status (2026-04-27)

**Total**: 142 papers -- 61 read, 23 skimmed, 53 unread, 5 cited, 0 dismissed.
**Most-referenced unread** (top 5):
| Citekey | Title | References in active project notes |
|---|---|---|
| kim2023ionic | ... | 7 |
| ... |

**Tags** (top 8):
- ion-gated-transistor (38), reservoir-computing (29), edl (22), ...

**Gaps** (cited in 10_Projects/{active-slug} but not in library): 3
- 10.1038/s41928-... -> suggested: /capture-paper 10.1038/s41928-...
- ...
```