Enables running code against tests in an isolated sandbox to obtain PASS/FAIL verdicts with signed, offline-checkable certificates, and generating verified code with attached certificates after execution against derived tests.
Enables an assistant to verify whether claims about AI systems hold against supplied evidence bytes, replay conformance vectors, and inspect related terms and artifact digests offline without fetching, writing, or sending data.
Enables deterministic verification of AI agent decisions and actions, providing PASS/FAIL/ABSTAIN verdicts with replayable proofs and an optional signed receipt ledger.
Enables agents to analyze software artifacts through 35 tools covering read-only reconnaissance, secret/string/symbol scanning, native and Dart AOT structure checks, exposure tiering, protection planning, attack-simulation design, and an append-only evidence ledger with mechanical verified gates that treat unresolved unknowns as blocking. Gated execution tools apply reviewed transforms only to a managed copy inside an authorized workspace, with one-time approval tokens, independent verification, before/after diffs, and rollback.
Enables coding agents to run bounded local profiling, tracing, and benchmarking analyses on explicit artifacts or commands, returning typed evidence with provenance, coverage, and limitations.
Exposes bounded local runtime-evidence tools that let coding agents analyze explicit artifact paths and capture explicit commands, coordinating profilers, benchmark tools, trace processors, and direct targets without any workspace or project configuration. Returns observed inputs, provenance, typed evidence, coverage, and limitations inline, with optional durable preservation and replay.
