PeriA builds infrastructure-layer tools that make AI agents faster, cheaper, and more reliable. Our first product, PRECC, cuts Claude API costs by up to 66% with a under 3ms hook — zero changes to your workflow.
Note: Real savings will depend heavily on how you use Claude Code and what kind of commands you're running.
Every product we ship targets a concrete inefficiency in how AI agents consume compute and APIs.
A pre-execution hook that intercepts, rewrites, and optimises every bash command Claude issues — before tokens are wasted.
Weekly digest of your AI API spend, breakdown by session, and personalised tuning recommendations. Automated, no dashboards.
Community-contributed rewrite rules and skill packs. Install, publish, and monetise skills that extend PRECC's capabilities.
PRECC registers as a Claude Code pre-execution hook. Every command passes through a staged pipeline in under 3ms — then continues as normal.
Read the docs →One curl command installs PRECC and registers the hook with Claude Code. No config required.
Claude issues a bash command. PRECC receives it before execution via the hook protocol.
6 skill stages run in parallel — context fix, RTK compression, jj translation, warning ranking, and more.
The rewritten command runs. Fewer tokens consumed. Fewer retries. Lower cost.
Free forever for community tier. No credit card. One command.
No data will be collected without your consent.