Releases: lroolle/wxclawbot-cli
Releases · lroolle/wxclawbot-cli
v0.5.2
v0.5.1
Structured transport error diagnostics
Transport-layer failures now include actionable fields in --json output:
{"ok":false,"error":"send failed: ...","errorKind":"timeout","retryable":true}Error kinds: timeout, dns, connection, network, tls, unknown.
Check retryable to decide whether to retry or fail fast.
Distributions
- npm:
npm install -g @claw-lab/wxclawbot-cli@0.5.1 - ClawHub:
clawhub install wxclawbot-send
v0.5.0
Proactive push fix
Messages now actually push to WeChat as notifications, instead of silently
sitting on the server until you open the chat.
What changed
- context_token support: reads openclaw-weixin's persisted context tokens
and includes them in every sendmessage request. This is what tells WeChat
to push the message as a notification. - One item per request: text and media are now sent as separate requests
(matching openclaw-weixin behavior), not batched into a single item_list.
Upgrade
npm install -g @claw-lab/wxclawbot-cli@0.5.0No config changes needed -- context tokens are auto-discovered from
openclaw-weixin's account store.
Distributions
- npm:
@claw-lab/wxclawbot-cli@0.5.0 - ClawHub:
clawhub install wxclawbot-send - GitHub: https://github.com/lroolle/wxclawbot-cli
v0.4.2
Patch release to sync all distributions.
- Account discovery docs (bot ID extracted at runtime, not hardcoded)
- Rate limit clarification (shared across ALL clients on same token)
- Install/update prompt in SKILL.md
Distributions:
- npm:
npm install -g @claw-lab/wxclawbot-cli@0.4.2 - ClawHub:
clawhub install wxclawbot-send
v0.4.1
Docs patch: install prompt now explains skill mechanism (SKILL.md placement + CLI dep + clawhub optional).
v0.4.0
Breaking: moved to @claw-lab/wxclawbot-cli
Package renamed from @herai/wxclawbot-cli to @claw-lab/wxclawbot-cli. The old package is deprecated on npm.
npm install -g @claw-lab/wxclawbot-cliChanges
- npm-first install -- install prompt uses
npm installinstead ofclawhub install(clawhub not default in openclaw) - @CLAW-Lab org -- package now under claw-lab npm org
- README rewrite -- zh-CN first, install as copy-paste prompt to agent, scenarios as natural language
Full Changelog
v0.3.0
What's Changed
Agent-first README
- README rewritten with AI agent prompt section first — setup check, send commands, output parsing, error handling
- Human docs and zh-CN below the fold
ClawHub Skill Publishing
- SKILL.md now includes
metadata.openclawblock (ClawHub-compatible) - Release workflow publishes to both npm and ClawHub
- Added
.clawhubignorefor clean skill bundles