Menu
Extract PDF text/metadata/page info using Sylphx PDF Reader MCP through a local CLI wrapper. Use when users want PDF extraction in pi without MCP integration.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Manage Robinhood account tasks including portfolio checks, quotes, watchlists, orders, options, scanners, earnings, and realized P&L using Robinhood MCP tools. Use for any user request involving Robinhood, brokerage accounts, stocks/options trading, watchlists, scans/screeners, portfolio value, orders, positions, or market data from Robinhood.
Administer Ronak's VPS at 5.78.176.171, including Docker Compose stacks, Caddy, the ronakpjain.com website, BlueMap, Minecraft/Purpur, DNS/routing checks, logs, firewall, users, SSH, and general server maintenance.
Executes implementation work in small validated increments and reports progress clearly. Use when carrying out an approved plan.
Performs discovery-first analysis of a codebase or system without making changes. Use to map architecture, behavior, and unknowns before planning.
Produces a concrete implementation plan with ordered steps, validation criteria, and rollback strategy. Use after exploration and before coding.
| name | pdf-reader-cli |
| description | Extract PDF text/metadata/page info using Sylphx PDF Reader MCP through a local CLI wrapper. Use when users want PDF extraction in pi without MCP integration. |
| allowed-tools | bash read |
Use this skill to read PDF files via @sylphx/pdf-reader-mcp in environments where pi does not use MCP directly.
SylphxAI/pdf-reader-mcp and wants equivalent capability in pi.No permanent install is required. The wrapper uses:
bunx @sylphx/pdf-reader-mcp
(First run may take longer due to package download.)
bun ~/.pi/agent/skills/pdf-reader-cli/scripts/read-pdf.mjs <path-or-url> [options]
--pages "1-3,8" page selection--no-text disable full text extraction--no-metadata disable metadata extraction--no-page-count disable page count--images include image info--tables include table detection--raw print raw MCP tools/call payload--timeout <ms> request timeout (default 120000)# Default: text + metadata + page count
bun ~/.pi/agent/skills/pdf-reader-cli/scripts/read-pdf.mjs ./docs/report.pdf
# Specific pages
bun ~/.pi/agent/skills/pdf-reader-cli/scripts/read-pdf.mjs ./docs/report.pdf --pages "1-5,10"
# URL source
bun ~/.pi/agent/skills/pdf-reader-cli/scripts/read-pdf.mjs "https://arxiv.org/pdf/2301.00001.pdf"
# Metadata only
bun ~/.pi/agent/skills/pdf-reader-cli/scripts/read-pdf.mjs ./docs/report.pdf --no-text
If this skill is triggered via /skill:pdf-reader-cli <args>, parse <args> as CLI arguments and run the command above.
After running, summarize key extracted content for the user (and include warnings/errors if extraction failed).