@agent-rewind/core
Advanced tools
Reversible execution for AI agents: whole-workspace checkpoint/rewind, a refuse-and-record effect barrier, and a tamper-evident SHA-256 evidence chain — pure logic over logs and traces.
The portable core of Agent Rewind — reversible execution for AI agents, as pure logic over logs and traces (never the filesystem or clock, so it runs unchanged across backend tiers).
Provides:
WorldBackend — the whole-workspace snapshot/restore interface, with a Tier-0 git implementation
(createGitBackend): append-only snapshot chain, atomic restore that rolls back on a mid-flight
failure, and secret/heavy-dir excludes.createEffectLedger) — admit-once / refuse-and-record over
(scope, effectKey) identity, so a spent external effect can't be re-fired across a rewind.createMemoryEvidenceLedger, computeEntryHash, verifyChain) —
append-only, per-scope, SHA-256-linked, fail-closed verification. The action vocabulary and authority
resolver are injected, so it carries no product-specific vocabulary.canonicalize), the engine (createEngine), a replay-savings sink, and
an AgentRewind-style recovery/failure-memory model.npm install @agent-rewind/core
Requires Node ≥ 20. Licence: FSL-1.1-ALv2. See the monorepo for architecture and the full product breakdown.
FAQs
Reversible execution for AI agents: whole-workspace checkpoint/rewind, a refuse-and-record effect barrier, and a tamper-evident SHA-256 evidence chain — pure logic over logs and traces.
The npm package @agent-rewind/core receives a total of 1 weekly downloads. As such, @agent-rewind/core popularity was classified as not popular.
We found that @agent-rewind/core demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Product
Socket Threat Intel now shows which malware campaigns were blocked or detected in your organization, with links to scans and Firewall events.
Security News
Fixes for eMudhra and Brotli are out. Getting them everywhere they’re needed means tracking down bundled libraries, runtimes, and certificate stores.
Threat Research
/Security News
Socket uncovered secret-stealing workflows that ran in LiteLLM’s GitHub repository 11 days before the supply chain attack on PyPI and Docker.