Skip to content

FAQ

Direct answers, in order.

An AGENTS.md where every enforceable claim carries an enforcement line (which gate, which toolchain, live status), is backed by a test that live-fires the negative example against the real rule, and whose body is generated and byte-gated against drift. See Executable AGENTS.md, defined for the full walkthrough.

No. The enforcement gates it generates are native toolchain gates (ESLint, husky, clippy, cargo-deny) that run deterministically, with no model call in the loop. $0 LLM calls in CI is enforced by a test in getff’s own repo, not a marketing claim.

Source-available under FSL-1.1-ALv2, not OSI open source. Every release converts to Apache-2.0 two years after it ships. We say this first, on purpose.

Packmind runs an LLM-in-the-loop standards platform with its own runtime and org-wide features. getff compiles conventions into native toolchain gates instead — no runtime of its own, no LLM call to enforce a rule. If you need an LLM-driven standards platform with organization-wide rollout features, that’s a real reason to pick Packmind instead.

agnix lints your agent configuration files. getff takes a different problem: making the claims inside those files executable, so a claim that stops being true fails a gate instead of silently going stale.

How is this different from CodeRabbit or Qodo?

Section titled “How is this different from CodeRabbit or Qodo?”

CodeRabbit and Qodo do LLM-powered code review, which is a genuinely useful complement, not a replacement. getff’s gates are deterministic and cost nothing per run, because they’re native toolchain rules (ESLint/husky, clippy/cargo-deny), not a model call.

The generated gates do, yes. Once getff compiles your conventions into ESLint rules, husky hooks, clippy lints, or cargo-deny policy, those gates live in your toolchain and run in any editor, any CI, with or without Claude Code.