promptready-mcp

Upload a local PDF/CSV and queue conversion (same credits as the web app). Uses **saved user settings** (get_convert_settings / set_convert_settings), not per-call engine flags. Factory default: paddle, tables on, images off. Args: path: Absolute or home-relative path to a .pdf or .csv file. wait: If true, poll until done and download markdown (can take many minutes). timeout_sec: Max wait when wait=true (default 3600). output_dir: Directory for downloaded .md when wait=true. Returns JSON with job info; if wait=true and completed, includes local md path. The download can lag completion by a few seconds; if the file still cannot be saved, the response says download="pending" — the conversion succeeded and credits were already spent, so call wait_and_download, not convert again.

Show saved convert defaults for this machine (not per-file options). Defaults live in ~/.config/promptready/settings.json. convert_pdf always uses these until you change them with set_convert_settings.

Get the calling user's PromptReady credit balance. Returns JSON: ok + credit_balance + email, or ok:false + error. Requires prior login (promptready-mcp-login / login tool) or env token.

Get conversion status — durable by log_id, session memory otherwise. With log_id (the UUID from convert_pdf's response): reads the DB-backed usage row (GET /users/usage) — survives backend restarts, redeploys and back-to-back converts. Status is the DB value (pending / completed / failed / cancelled). Without log_id: reads the session status (GET /convert/status); if the server no longer knows the session job ("not_found" — restart, scale-to-zero), falls back to the most recent usage row and says which file it picked.

List this account's recent conversions (DB-backed, restart-proof). Each entry: log_id, file_name, page_count, credits, status (pending/completed/failed/cancelled), created_at, expires_at, engine_used, and downloadable (completed ∧ result retained — false once the 3-hour window passed). Use a log_id with wait_and_download to fetch any downloadable result, even after a server restart.

Open the browser Google login page and save tokens for this machine. Starts a local callback on http://127.0.0.1:18765/callback, opens PromptReady Google OAuth, stores credentials under ~/.config/promptready/credentials.json (mode 600), and applies them to this process. Users should run this once (or use CLI: promptready-mcp-login) instead of pasting tokens manually. After sign-in the browser returns to the local callback URL; no Supabase configuration is required.

Remove saved MCP credentials from this machine and clear env tokens.

Update saved convert defaults (persists for future convert_pdf calls). Only pass fields you want to change. Example: set engine to paddle once, then every convert uses that until changed again. Args: engine: paddle (PaddleOCR-VL-1.6) | deepseek_ocr (DeepSeek-OCR) | glm_ocr (GLM-OCR) include_tables: keep markdown tables include_images: include image/visual extraction (heavier) remove_references: strip references section when supported

Wait for a conversion to finish, then save its markdown. Durable path (0.4.0): pass log_id (UUID from convert_pdf's response or list_conversions) — status comes from the DB usage row and the file from GET /convert/download/{log_id}, so the result survives backend restarts and later converts. 410/404 there mean the 3-hour retention window closed: the stored result is gone and converting again (fresh credits) is the only way to get it. Without log_id: polls the session status as before (free reads); if the server reports not_found, falls back to the most recent usage row and says which file was picked (plus a warning when it does not match source_path). Args: timeout_sec: Max seconds to wait (OCR can take many minutes). poll_interval_sec: Sleep between polls (default 3). output_dir: Where to write the .md file. source_path: Optional original file path (fallback filename check). log_id: Optional conversion id (UUID) — preferred, restart-proof.