← 返回 Skills 市场
shan8851

OpenClaw Blackbox

作者 shan8851 · GitHub ↗ · v0.1.0 · MIT-0
cross-platform ✓ 安全检测通过
59
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install openclaw-blackbox
功能描述
OpenClaw Blackbox CLI — inspect failed, stalled, expensive, or weird OpenClaw agent runs from local disk evidence; list sessions and failures; find the sessi...
使用说明 (SKILL.md)

OpenClaw Blackbox

Use blackbox to inspect local OpenClaw run evidence and produce deterministic debugging reports.

Setup

  • npm install -g @shan8851/blackbox
  • Requires local OpenClaw state; by default it reads ~/.openclaw
  • To inspect another OpenClaw home: OPENCLAW_HOME=/path/to/.openclaw blackbox doctor

Health Check

  • Check visible OpenClaw evidence: blackbox doctor
  • JSON health check: blackbox doctor --json

List Evidence

  • Recent failed or suspicious cron runs: blackbox list failures --limit 10
  • Filter by label: blackbox list failures --label context_overflow
  • Emit runnable inspect commands: blackbox list failures --commands
  • Local sessions across agents: blackbox list sessions --agent all --limit 20
  • Sessions with specific evidence: blackbox list sessions --evidence checkpoint_present

Inspect Runs

  • Latest failed run for a cron job: blackbox inspect --cron-job "Example nightly job" --latest-error
  • Known session: blackbox inspect --session-id \x3Csession-id>
  • Full Markdown report: blackbox inspect --session-id \x3Csession-id> --view full --out reports/run.md
  • JSON report: blackbox inspect --session-id \x3Csession-id> --json-out reports/run.json
  • HTML report: blackbox inspect --session-id \x3Csession-id> --html-out reports/run.html --open

Find Requests

  • Find the session behind a user request: blackbox find request --query "summarise the failed run" --agent all
  • Find by message id: blackbox find request --message-id \x3Cmessage-id> --agent main
  • Emit inspect commands only: blackbox find request --query "failed to send" --commands

Output

  • inspect always prints a terminal snapshot
  • Markdown defaults to reports/\x3Crun>.md unless --out is set
  • Set BLACKBOX_REPORT_DIR to change the default report directory
  • Discovery commands support --json for machine-readable output
  • --view simple is the default triage report; --view full includes fuller evidence and timeline details

Agent Notes

  • Prefer Blackbox before guessing why an OpenClaw run failed if local evidence exists
  • find request / inspect --query are conservative; if a query maps to multiple sessions, inspect by explicit --session-id
  • Reports can include prompts, tool arguments/results, URLs, and local paths; review before sharing externally
  • Blackbox is disk-first and should not make hidden LLM calls by default
  • Failure labels include context_overflow, message_delivery_failed, timeout, session_lock_timeout, gateway_restart_interrupted, model/auth labels, tool failure labels, and unknown_error
  • Evidence labels include missing_transcript, transcript_deleted, transcript_reset, missing_trajectory, checkpoint_present, and session_write_in_progress

Notes

  • Exit codes: 0 success, 1 unexpected failure, 2 usage/invalid argument, 3 not found/missing OpenClaw state, 4 ambiguous request match
  • Package name: @shan8851/blackbox; binary name: blackbox; skill/ClawHub slug: openclaw-blackbox
安全使用建议
This skill appears coherent and purpose-aligned for debugging OpenClaw runs. Before installing, make sure you trust the @shan8851/blackbox npm package, because its code was not included in the submitted artifacts. When using it, remember that reports may contain sensitive local run details; review and redact them before sharing.
功能分析
Type: OpenClaw Skill Name: openclaw-blackbox Version: 0.1.0 The skill provides a CLI tool (`blackbox`) for debugging local OpenClaw agent runs by inspecting the `~/.openclaw` directory. The instructions and command set in SKILL.md are consistent with the stated purpose of triage and reporting. It includes a security warning regarding the sensitivity of generated reports, and no indicators of malicious intent or high-risk behavior were found in the provided files.
能力评估
ℹ Purpose & Capability
The skill’s purpose is coherent: it uses the Blackbox CLI to inspect local OpenClaw run evidence and generate debugging reports. That purpose naturally involves reading potentially sensitive run transcripts and evidence.
✓ Instruction Scope
The instructions are framed as user-invoked diagnostic commands, include ambiguity handling guidance, and warn that reports should be reviewed before external sharing.
ℹ Install Mechanism
The skill requires installing and running the external npm package @shan8851/blackbox as the blackbox CLI. No package source files were included in the submitted artifact set, so package behavior is not independently reviewable here.
ℹ Credentials
Reading ~/.openclaw is proportionate for OpenClaw run debugging, but it can expose private local session evidence across agents if broad options such as --agent all or full reports are used.
ℹ Persistence & Privilege
The skill can create persistent local Markdown, JSON, or HTML reports. There is no evidence of background persistence, privilege escalation, or automatic external sharing.
如何使用
  1. 确保已安装 OpenClaw（本地或 Docker 部署）
  2. 在对话框中输入安装命令：/install openclaw-blackbox
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /openclaw-blackbox 触发
  4. 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.1.0
Initial OpenClaw Blackbox companion skill
元数据
Slug openclaw-blackbox
版本 0.1.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

OpenClaw Blackbox 是什么？

OpenClaw Blackbox CLI — inspect failed, stalled, expensive, or weird OpenClaw agent runs from local disk evidence; list sessions and failures; find the sessi... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 59 次。

如何安装 OpenClaw Blackbox？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install openclaw-blackbox」即可一键安装，无需额外配置。

OpenClaw Blackbox 是免费的吗？

是的，OpenClaw Blackbox 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。

OpenClaw Blackbox 支持哪些平台？

OpenClaw Blackbox 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。

谁开发了 OpenClaw Blackbox？

由 shan8851（@shan8851）开发并维护，当前版本 v0.1.0。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191113 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论