Skip to main content
RTILA Marketplace

The best ai automation tools in one place

AI automation tools: workflows that use AI completion steps for classification, extraction and drafting inside a run, with provider and key requirements published. 0 bots in this category.

AI listings use the engine's completion command inside an otherwise ordinary automation run: classifying collected rows, extracting structured fields from messy text, drafting the first pass of a summary from data the run gathered. The AI step is a stage in a deterministic pipeline - and the listings say so plainly, including where output varies between runs.

Expect the provider and model requirements published on the Input tab, your own API key prompted at run time and never stored in the configuration, and an honest non-determinism note: a completion stage does not reproduce byte-identical output, and no listing claims it does.

Every listing ships the same four launchers - a Windows .exe and .sh scripts for macOS Apple Silicon, macOS Intel, and Linux - delivered as a single raw file with no archive or extraction step. Restore the executable bit on the shell-script platforms, and clear the macOS Gatekeeper quarantine attribute before the first run. AI stages add per-row provider cost to a run: the listing's row caps and prompts exist so a scheduled job cannot spend your key's budget unattended.

Bots in this category

No bots in this category yet

Check back soon, or browse another category.

AI bots: frequently asked questions

What do AI tools automate here?
The judgement-adjacent middle of a pipeline: classify these collected rows, extract these fields from this text, draft a summary of this dataset. Collection before it and decisions after it stay deterministic and yours.
Do I need my own AI provider key?
Yes, wherever a listing declares an AI step: the key is prompted at run time as a secret input and is never embedded in the published configuration.
What do AI runs cost?
The launcher is free; the provider meters the completion calls against your own key. Row caps in the run prompt bound what one execution can spend.
Is the output deterministic?
The pipeline around the AI step is; the completion itself is not, and can vary between runs on identical input. Listings that include an AI stage state this rather than hide it.