Overleaf logo

Overleaf

Community
aloth
overleaf

Sync and manage Overleaf LaTeX projects from the command line. Pull projects locally, push changes back, compile PDFs, and download compile outputs like .bbl files for arXiv submissions. Use when working with LaTeX, Overleaf, academic papers, or arXiv.

Overview

Publisheraloth
Repositoryolcli
Skill nameoverleaf
Stars
220
Forks
29
Bundled files
51
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 51 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by aloth on GitHub. Read the source before you install it.

Installation

Install the Overleaf AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/aloth/olcli/tree/main
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/aloth/olcli.git \
  .claude/skills/overleaf
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Overleaf in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Overleaf on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Overleaf is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Overleaf Skill

Manage Overleaf LaTeX projects via the olcli CLI, native git remote, or MCP server.

When to Use Which Mode

ModeBest forHow
CLI (olcli)Interactive workflows, sync, compile, arXiv prepolcli pull/push/sync/pdf
Git remoteVersion control, commits, diffs, CI/CD pipelinesgit clone overleaf::… then standard git
MCP serverAI agents with MCP support (Claude, Cursor, Windsurf)Connect via olcli-mcp stdio transport

Use CLI when you need bidirectional sync with conflict detection, compilation, or comment management. Use Git remote when you want proper git history, branches, and standard git push/pull. Use MCP when an AI agent has native MCP support and doesn't need to shell out.

Installation

bash
# Homebrew (recommended)
brew tap aloth/tap && brew install olcli

# npm
npm install -g @aloth/olcli

Authentication

Get your session cookie from Overleaf:

  1. Log into overleaf.com
  2. Open DevTools (F12) → Application → Cookies
  3. Copy the value of overleaf_session2
bash
olcli auth --cookie "YOUR_SESSION_COOKIE"

Verify with:

bash
olcli whoami

Debug authentication issues:

bash
olcli check

Clear stored credentials:

bash
olcli logout

Clears the global config and the .olauth file in the current directory, and reports each. Environment variables cannot be unset by a child process, so OVERLEAF_SESSION and OVERLEAF_EMAIL/OVERLEAF_PASSWORD are reported rather than silently ignored — they outrank anything on disk.

For unattended use, prefer OVERLEAF_EMAIL/OVERLEAF_PASSWORD over olcli auth --password: every command reads them, and nothing is written to disk or to shell history.

Self-hosted Overleaf

bash
olcli config set-url https://overleaf.yourcompany.com
olcli config set-cookie-name overleaf.sid   # if different from default
olcli auth --cookie "YOUR_COOKIE"

Or pass per-command: olcli --base-url https://overleaf.yourcompany.com list

Git Remote Helper

Use Overleaf projects as native git remotes. No wrapper scripts needed.

bash
# Clone
git clone overleaf::https://www.overleaf.com/project/<id>
cd <project>

# Edit, commit, push — standard git workflow
vim main.tex
git add . && git commit -m "update introduction"
git push

# Pull latest from Overleaf
git pull

Authentication: reads OVERLEAF_SESSION env var, ~/.olauth file, or stored config (same as CLI).

For self-hosted instances, just use your instance URL:

bash
git clone overleaf::https://overleaf.yourcompany.com/project/<id>

Debug with: GIT_REMOTE_OVERLEAF_DEBUG=1 git push

MCP Server

Built-in Model Context Protocol server for AI assistant integration.

bash
# Run standalone
olcli-mcp

# Or via npx
npx @aloth/olcli-mcp

Available MCP tools: list_projects, get_project_info, pull_project, push_file, compile, download_pdf, list_comments, get_entities, download_file, add_comment, reply_to_comment, resolve_comment, delete_entity, rename_entity, rename_project, plan_project_renames, compile_with_outputs, diff_project, create_project.

compile, download_pdf and compile_with_outputs accept an optional resource_path to compile a specific root document.

diff_project is the MCP counterpart of olcli diff: read-only, fetches the remote fresh on every call, and returns one entry per changed file with path, status, binary and a unified patch. Pass name_only to drop the patch text. plan_project_renames previews bulk renames and never applies them.

Auth: set OVERLEAF_SESSION env var in MCP config, or use stored credentials from olcli auth.

Common Workflows

Pull a project to work locally

bash
olcli pull "My Paper"
cd My_Paper/

Create a project

bash
olcli project create "My Paper"
olcli project create "Example Paper" --template example

Edit and sync changes

bash
# After editing files locally
olcli push              # Upload changes only
olcli sync              # Bidirectional sync (pull + push, propagates local deletions)
olcli sync --no-delete  # Sync without propagating local deletions to remote

Review changes before pushing

bash
olcli diff                 # unified diff of every changed file
olcli diff --name-only     # changed paths only
olcli diff --file main.tex # a single file
olcli diff --exit-code     # CI gate: 0 same, 1 differs, 2 failed

The remote side is fetched fresh each run, so this shows what a subsequent push would overwrite — not a comparison against the last pull. a/ is the remote, b/ is local. Binary files are reported as differing without a patch.

--exit-code uses diff(1)'s statuses so the command can gate a pipeline: 0 nothing differs, 1 something does, 2 the run itself failed. The last one matters — without it a job cannot tell a changed file from an expired session. Failures stay 1 when the flag is absent, so existing scripts are unaffected.

Delete or rename remote files

bash
olcli delete chapters/old.tex          # remove a file from the project
olcli rm figures/old.pdf               # alias
olcli rename old.tex new.tex           # rename a file
olcli mv chapters/draft.tex chapters/intro.tex   # alias

Inspect ignore rules

bash
olcli ignored              # list active patterns (built-ins + .olignore + .olignore.local)
olcli push --show-ignored  # see what was filtered on this run
olcli sync --no-ignore     # escape hatch: upload everything

Compile and download PDF

bash
olcli pdf                      # Compile and download
olcli pdf -o paper.pdf         # Custom output name
olcli pdf -r chapters/intro.tex  # Compile a specific root document
olcli compile                  # Just compile (no download)
olcli compile -r appendix.tex  # Compile a specific root document without downloading

-r, --resource <path> works on compile, pdf, and output: it compiles the given .tex file as the root document. Useful when a project contains several documents.

Download .bbl for arXiv submission

bash
olcli output bbl               # Download compiled .bbl
olcli output bbl -o main.bbl   # Custom filename
olcli output bbl -r appendix.tex
olcli output --list            # List all available outputs

Upload figures or assets

bash
olcli upload figure1.png "My Paper"          # Upload to project root
olcli upload diagram.pdf                      # Auto-detect project from .olcli.json
olcli upload figures/diagram.png              # Relative path is preserved remotely
olcli upload /tmp/build/diagram.png           # Absolute path lands in the project root
olcli upload /tmp/build/diagram.png --to figures/diagram.png   # Explicit destination

Remote path rules: a relative local path keeps its directory part, an absolute local path collapses to its basename, and --to overrides both.

Download specific files

bash
olcli download main.tex "My Paper"           # Download single file
olcli zip "My Paper"                          # Download entire project as zip

Review comments

bash
olcli comments list                          # List all comments (current project)
olcli comments list --status open            # Filter by status (open/resolved/all)
olcli comments list --context                # Include surrounding text
olcli comments add main.tex "Fix this citation" --from 10 --to 15  # Add comment
olcli comments reply <thread-id> "Done!"     # Reply to thread
olcli comments resolve <thread-id>           # Mark as resolved
olcli comments reopen <thread-id>            # Reopen a resolved thread
olcli comments delete <thread-id>            # Delete entire thread

arXiv Submission Workflow

Complete workflow for preparing an arXiv submission:

bash
# 1. Pull your project
olcli pull "Research Paper"
cd Research_Paper

# 2. Compile to ensure everything builds
olcli compile

# 3. Download the .bbl file (arXiv requires .bbl, not .bib)
olcli output bbl -o main.bbl

# 4. Download any other needed outputs
olcli output aux -o main.aux    # If needed

# 5. Package for submission
zip arxiv.zip *.tex main.bbl figures/*.pdf

# 6. Verify the package compiles locally (optional)
# Then upload arxiv.zip to arxiv.org

Commands Reference

CommandDescription
olcli auth --cookie <value>Authenticate with session cookie
olcli auth --email <e>Authenticate with password, prompted (self-hosted)
olcli whoamiCheck authentication status
olcli logoutClear the global config and the local .olauth
olcli checkShow config paths and credential sources
olcli listList all projects
olcli project create <name>Create a blank or example project
olcli info [project]Show project details
olcli pull [project] [dir]Download project files
olcli push [dir]Upload local changes
olcli sync [dir]Bidirectional sync
olcli diff [project] [dir]Content-level diff of local files vs. the live remote
olcli upload <file> [project]Upload a single file (--to <path> sets the remote destination)
olcli download <file> [project]Download a single file
olcli delete <file> [project]Delete a remote file or folder (alias: rm)
olcli rename <old> <new> [project]Rename a remote file or folder (alias: mv)
olcli ignored [dir]List active ignore patterns
olcli zip [project]Download as zip archive
olcli compile [project]Trigger compilation
olcli pdf [project]Compile and download PDF
olcli output [type]Download compile outputs
olcli comments list [project]List review comments
olcli comments add <file> <msg>Add a comment
olcli comments reply <id> <body>Reply to a thread
olcli comments resolve <id>Resolve a thread
olcli comments reopen <id>Reopen a thread
olcli comments delete <id>Delete a thread
olcli config set-url <url>Set self-hosted base URL
olcli config get-urlShow the configured base URL
olcli config set-cookie-name <name>Set cookie name
olcli config get-cookie-nameShow the configured cookie name
olcli config set-timeout <ms>Set HTTP timeout
olcli config get-timeoutShow the configured HTTP timeout
olcli project rename <old> <new>Rename a project
olcli project rename-bulkRename many projects by pattern (dry-run unless --apply)

Tips

  • Auto-detect project: Run commands from a synced directory (contains .olcli.json) to skip the project argument
  • Dry run: Use olcli push --dry-run or olcli sync --dry-run to preview before applying
  • Preview content: push --dry-run lists files by modification time; olcli diff compares actual contents, so the two lists can differ
  • CI gate: olcli diff --exit-code exits 1 when anything differs and 2 when the run failed, so a pipeline can distinguish drift from breakage
  • Force overwrite: Use olcli pull --force to overwrite local changes
  • Two-way deletes: olcli sync propagates local deletions to the remote; use --no-delete to opt out per run
  • Build artifacts: .aux, .bbl, .log, .synctex.gz etc. are filtered by default. Add custom patterns to a .olignore file (gitignore-style)
  • PDF rule: thesis.pdf next to thesis.tex is auto-ignored; standalone figures/diagram.pdf is preserved
  • Project ID: You can use project ID instead of name (24-char hex from URL)
  • Debug auth: Run olcli check to see where credentials are loaded from
  • Timeout: olcli --timeout 60000 pull "Big Project" or olcli config set-timeout 60000

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Overleaf AI skill do?

Sync and manage Overleaf LaTeX projects from the command line. Pull projects locally, push changes back, compile PDFs, and download compile outputs like .bbl files for arXiv submissions. Use when working with LaTeX, Overleaf, academic papers, or arXiv.

Why use Overleaf on TypingMind?

Because you install it once and use it with any model. Overleaf is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Overleaf in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/aloth/olcli/tree/main. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Overleaf?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Overleaf?

As many as you like. As long as a model supports skills, you can use Overleaf with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Overleaf AI skill free?

Yes. It is published on GitHub by aloth under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇