review-overleaf
The review-overleaf skill fetches unresolved review threads from an Overleaf LaTeX project, matches comments to their locations in a local git-tracked repository, and applies proposed edits based on reviewer feedback. Use this skill when you need to incorporate Overleaf comments into your local LaTeX files without manually finding and applying each change.
git clone --depth 1 https://github.com/fcakyon/claude-codex-settings /tmp/review-overleaf && cp -r /tmp/review-overleaf/plugins/overleaf-skills/skills/review-overleaf ~/.claude/skills/review-overleafSKILL.md
# Review Overleaf
Pull unresolved review threads from an Overleaf project, locate each in the local repo, propose edits the user reviews, and apply them. Does not push back to Overleaf — the web UI is still the place to mark threads resolved after the user verifies the local change.
## Step 1: Resolve project ID
Accept any of:
- Raw 24-char hex ID (`^[0-9a-f]{24}$`)
- Full URL like `https://www.overleaf.com/project/<id>` — extract via regex
- Project name in quotes — call:
```bash
python3 ${CLAUDE_PLUGIN_ROOT}/scripts/overleaf_reviews.py --list-projects
```
Each line is `<24-hex> <accessLevel> <name>`. Fuzzy-match the name. If multiple match, AskUserQuestion to disambiguate.
If the user has not given a project ID at all, run `--list-projects` and show the user the table so they can pick one.
## Step 2: Cookie precheck
If `~/.claude/overleaf-skills/cookie` does not exist, tell the user to "set up overleaf" (which triggers the `setup` skill) and stop.
If it exists but auth fails (the next script call returns a refresh hint), point at the same setup skill and stop.
## Step 3: Access-level guard
If the project ID came from name resolution, check the matched entry's `accessLevel`. If `readOnly`, warn:
> Project is read-only. You can edit local files but cannot sync them back to Overleaf via the web UI; you would need to copy-paste manually. Continue?
For `owner`, `readAndWrite`, or `review`, proceed silently.
## Step 4: Fetch comments
```bash
python3 ${CLAUDE_PLUGIN_ROOT}/scripts/overleaf_reviews.py --json --repo . < project-id > --unresolved-only
```
Parses to a JSON array. Each record has: `thread_id`, `doc_id`, `snippet`, `offset`, `resolved`, `messages` (list of `{user, ts, content}`), `file`, `line`. The `file` and `line` are null when the snippet did not match any local `*.tex`.
## Step 5: Edit loop
For each record where `file` is non-null:
1. `Read` the file at `line` ± 5 lines for context.
2. Summarize the comment thread in one sentence: who said what.
3. Propose a concrete edit grounded in the comment text. If the comment is a question (e.g., "did you mean X?"), the edit should answer it in the prose; if it's a correction (e.g., "wrong notation"), the edit should make the correction.
4. Apply via `Edit`. Do not batch — one comment, one edit, then move on.
Skip records where `file` is null (snippet did not map). Collect them in an `<unmapped>` list.
## Step 6: Wrap up
After the loop:
1. Run `git diff` and show the user the full set of changes.
2. List the `<unmapped>` records (if any) with `doc_id` + snippet preview, so the user can hunt them manually.
3. Suggest a commit message in the style: `address overleaf review: <one-line summary of the changes>`. Do NOT auto-commit. The user can run `/github-dev:commit-staged` or commit by hand.
4. Remind the user to mark the addressed threads as resolved in the Overleaf web UI once they verify the local diff.
## Non-goals
- No re-upload to Overleaf (no write API used).
- No thread resolution in Overleaf (web UI only).
- No `.bib` edits unless a comment specifically targets a citation.
- No multi-line snippets that span paragraph breaks — those land in `<unmapped>` for v1; the snippet matcher is line-anchored.
## Pairs naturally with
- `github-dev` — `/github-dev:commit-staged` after the edit loop, `/github-dev:create-pr` if the project is collaborative.Agent-browser usage guide. Read this before running any agent-browser commands. Covers the snapshot-and-ref workflow, navigating pages, interacting with elements (click, fill, type, select), extracting text and data, taking screenshots, managing tabs, handling forms and auth, waiting for content, running multiple browser sessions in parallel, and troubleshooting common failures. Use when the user asks to interact with a website, fill a form, click something, extract data, take a screenshot, log into a site, test a web app, or automate any browser task.
Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include "automate Slack app", "control VS Code", "interact with Discord app", "test this Electron app", "connect to desktop app", or any task requiring automation of a native Electron application.
Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.
Use when tasks involve reading, creating, or reviewing PDF files where rendering and layout matter; prefer visual checks by rendering pages (Poppler) and use Python tools such as `reportlab`, `pdfplumber`, and `pypdf` for generation and extraction.
Use this skill any time a .pptx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates, layouts, speaker notes, or comments. Trigger whenever the user mentions \"deck,\" \"slides,\" \"presentation,\" or references a .pptx filename, regardless of what they plan to do with the content afterward. If a .pptx file needs to be opened, created, or touched, use this skill.
Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like \"the xlsx in my downloads\") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.
This skill should be used when user asks to "query Azure resources", "list storage accounts", "manage Key Vault secrets", "work with Cosmos DB", "check AKS clusters", "use Azure MCP", or interact with any Azure service.
This skill should be used when user encounters "Tavily MCP error", "Tavily API key invalid", "web search not working", "Tavily failed", or needs help configuring Tavily integration.