Scriptology Quick Start Guide

From Dashboard → Scriptology → Generate, describe what the script needs to do in plain English — for example, "watch a folder for new CSV files, validate the schema, and upload valid rows to Postgres in batches of 500." Choose a target language/platform, and Scriptology generates a complete, runnable script, not a snippet.

• →Be specific about inputs, outputs, and error handling in your description — "upload to Postgres" generates something usable; "upload to Postgres, skip rows missing an email, and log skipped rows to a file" generates something you can actually run in production
• →Every generated script runs through the same automated trust check used across the catalogue (see step 5) before you're shown the result

For anything that's genuinely more than one file — Dashboard → Scriptology → Packs — describe the whole thing at once, e.g. "a Cloudflare Worker API with JWT auth, rate limiting, a D1-backed users table, and CORS handling." A pack generates every file together, with imports and references between files kept consistent, unlike generating each file separately and wiring them up by hand.

• →Packs are the right tool the moment "script" stops being singular — an API needs routes, a schema, and middleware as consistent files, not three unrelated generations
• →Your own saved packs live under My Packs, separate from the public Packs catalogue you're browsing/generating into here

Build has four related actions for code you already have, not just fresh generation: Modify (change specific behavior while keeping the rest intact), Improve (readability, performance, or error handling passes), Convert (translate between languages/frameworks), and Explain (a plain-English walkthrough of what a piece of code actually does, useful for reviewing something before you run it).

• →Explain is worth using on any script you didn't write yourself before running it, generated by Scriptology or not
• →Convert preserves behavior, not just syntax — it's a genuine translation, not a find-and-replace

Validate checks a script for syntax correctness and obvious runtime issues without executing it. Audit goes further — a structured review covering security patterns, dependency risk, and code quality. Run both before Run (step 6), especially on anything Scriptology didn't generate itself.

• →Validate is fast enough to run after every edit; Audit is worth running once before anything goes into production use
• →Neither of these executes your code — they're static analysis, safe to run on anything

Library (Dashboard → Scriptology → Library) is your own saved scripts. Discover (the external link in the sidebar, at muqira.com/scriptology) is the full public catalogue — every published asset shows a trust score built from five real automated checks: syntax correctness, hardcoded-secret detection, dependency risk, security-pattern scanning, and freshness (how recently it was verified). It's not a popularity or star rating — it's a direct signal of what the automated checks actually found.

• →A lower trust score isn't automatically "don't use this" — open the asset and read which specific check flagged something; a freshness flag on an otherwise-clean script is very different from a secrets flag
• →Human-reviewed assets (marked explicitly) have had a person confirm the automated findings, not just the algorithm alone

Run (Dashboard → Scriptology → Run) executes a script through Muqira's sandboxed execution environment on Professional and Agency plans — useful for a quick one-off run without setting up your own environment first. This is separate from a Muqira Flow workflow step, which is for scripts that need to run automatically on a schedule or trigger, not on demand.

• →Sandbox execution has its own resource/time limits shown before you run — check them if your script does anything long-running
• →For anything that needs to run repeatedly or automatically, build it into a Muqira Flow workflow instead of re-running it manually here each time