npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@sovereignclaw/mesh

v0.2.0

Published

Multi-agent coordination for SovereignClaw — bus on 0G Log, patterns, replay.

Readme

@sovereignclaw/mesh

Multi-agent coordination for SovereignClaw. One Bus (append-only event log over any MemoryProvider), one Mesh (Bus + agent registry), and planExecuteCritique, the default planner → executor → critic orchestration pattern. Everything the agents say to each other is a typed, sequenced, optionally-encrypted event on the log.

Install

pnpm add @sovereignclaw/mesh @sovereignclaw/core @sovereignclaw/memory

10-line quickstart

import { Agent, sealed0GInference } from '@sovereignclaw/core';
import { InMemory } from '@sovereignclaw/memory';
import { Mesh, planExecuteCritique } from '@sovereignclaw/mesh';

const inf = () =>
  sealed0GInference({
    /* model, apiKey, baseUrl, verifiable: true */
  });
const mesh = new Mesh({ meshId: 'm1', provider: InMemory({ namespace: 'm1-bus' }) });
mesh
  .register(new Agent({ role: 'planner', systemPrompt: 'You plan.', inference: inf() }))
  .register(new Agent({ role: 'executor', systemPrompt: 'You execute.', inference: inf() }))
  .register(new Agent({ role: 'critic', systemPrompt: 'JSON only.', inference: inf() }));
const r = await planExecuteCritique({
  mesh,
  planner: mesh.get('planner')!,
  executors: [mesh.get('executor')!],
  critic: mesh.get('critic')!,
  task: 'Name the Transformer authors. One sentence.',
  maxRounds: 2,
  acceptThreshold: 0.7,
});

API

| Export | Kind | Purpose | | --------------------------- | ----- | -------------------------------------------------------------------------------------------------------- | | Bus | class | Append-only log: append(event), replay(fromSeq?), on(handler). Thin wrapper on a MemoryProvider. | | Mesh | class | Bus + agent registry. register / get / on / close. | | planExecuteCritique(opts) | fn | Planner → executor(s) → critic loop with acceptance threshold. | | BusEvent<P> | type | { type, fromAgent, seq, timestamp, parentSeq?, meshId, payload }. | | BusEventTypes | enum | TaskCreated / PlanCreated / ExecutionStarted / ExecutionComplete / CritiqueCreated / TaskComplete. | | SeqCounter | class | Monotonic sequence generator. Used by Bus internally. | | eventKey / seqFromKey | fn | Canonical, lex-sortable key encoding for bus events. |

Errors

All extend MeshError:

| Error | When | | ------------------------ | ----------------------------------------------------------- | | BusAppendError | Underlying provider set failed. | | BusReplayError | Underlying provider list or get failed. | | MeshClosedError | Operation after mesh.close(). | | PatternError | Parent class for pattern-specific failures. | | EmptyAgentOutputError | An agent returned null/empty text during the pattern. | | MaxRoundsExceededError | planExecuteCritique ran out of rounds without acceptance. | | CritiqueParseError | Critic output did not parse to {score, accept, ...}. |

Bus guarantees

  • Single-writer ordering. Inside one Mesh instance, sequence numbers are strictly monotonic. Cross-process replay with fan-in writers is explicitly deferred — see docs/dev-log.md Phase 5 notes.
  • Encryption-agnostic. The bus stores raw bytes; wrap the provider with encrypted(...) and the whole log is AES-GCM-sealed to a KEK derived from an EOA signer.
  • Deterministic replay. replay() returns events in seq order with stable keys so tests and audits can diff the log across runs.

Patterns

planExecuteCritique is the one shipped v0 pattern. Debate and hierarchical patterns are deferred to a future phase — see docs/dev-log.md. You can build your own pattern by composing mesh.bus.append(...) and agent.run(...) directly; there is nothing privileged about planExecuteCritique.

Further reading

License

MIT — see the repo root.