← 返回 Skills 市场
Pulse Board
作者
LittleJakub
· GitHub ↗
· v1.1.3
· MIT-0
405
总下载
0
收藏
0
当前安装
6
版本数
在 OpenClaw 中安装
/install pulse-board
功能描述
Universal operational digest for agent skill stacks. Every scheduled skill logs its outcome with log-append.sh. Twice daily, digest-agent.sh reads the log, c...
安全使用建议
This skill appears to do what it says. Before installing: (1) review the secrets env file it will source and ensure it contains only credentials you are comfortable sourcing into cron jobs; (2) verify the OpenClaw agent you configure uses a local LLM (e.g., Ollama) if you want to keep raw logs on-host — otherwise the raw pending.log will be included in the prompt and may be sent to a remote LLM; (3) ensure the jobs you plug do not print secrets to stdout/stderr (Pulse Board cannot prevent wrapped jobs from echoing secrets into their detail logs); (4) confirm the exact crontab changes when prompted by install.sh. If privacy is critical, either use the mechanical fallback only, run the digest agent with a local-only OpenClaw setup, or avoid enabling LLM composition.
功能分析
Type: OpenClaw Skill
Name: pulse-board
Version: 1.1.3
The Pulse Board skill provides operational monitoring by wrapping cron jobs and aggregating logs into digests. It exhibits high-risk behaviors including automated crontab modification (in install.sh and plug.sh) and the sourcing of a secrets environment file (~/.openclaw/shared/secrets/openclaw-secrets.env) within those cron jobs. Additionally, digest-agent.sh transmits raw log content to an LLM agent for summarization, and deliver.sh sends data to external Telegram or Discord endpoints. While these capabilities are aligned with the stated purpose and the installer includes explicit user confirmation steps and privacy warnings, the combination of crontab manipulation, secrets handling, and data exfiltration to external APIs warrants a suspicious classification.
能力评估
Purpose & Capability
Name/description match the implementation: scripts create ~/.pulse-board, append per-job lines, compose a digest via openclaw agent (or mechanical fallback), deliver to Telegram/Discord/log, and provide plug/unplug/install helpers. Required binaries and metadata correspond to the behavior.
Instruction Scope
Instructions and scripts do exactly what's described. They explicitly source a user-provided secrets env (if present) into cron-wrapped commands and pass the raw pending.log to the configured OpenClaw agent as prompt context — this is intentional and documented in multiple places. That behavior enlarges the privacy surface (logs may be transmitted off-host if the agent uses a remote LLM).
Install Mechanism
No remote download/install is performed — installer is an included script that writes files under the user's home and updates user crontab (via python3). No network installs or opaque external archives are used.
Credentials
No unexpected credentials or unrelated environment variables are required. Optional env vars and the secrets env file are used only to source delivery credentials and (optionally) LLM/workspace keys; install.sh only appends keys to the secrets env with explicit consent. The scope of env access is proportional to the stated functionality.
Persistence & Privilege
The skill does not request always:true and does not modify other skills' configurations. It writes only to ~/.pulse-board and the user crontab (user-level). Cron entries and registry files are created/removed via provided scripts with confirmation prompts.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install pulse-board - 安装完成后，直接呼叫该 Skill 的名称或使用
/pulse-board触发 - 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
v1.1.3
- `digest-agent.sh` and `deliver.sh` rewritten for clarity and compactness — same behaviour, less code. Single-letter internal helpers, config reads collapsed to one line each, mechanical fallback simplified.
- `deliver.sh` Telegram payload now has `parse_mode: Markdown` baked in — header `*Pulse Board Digest*` renders bold; LLM body stays plain text.
- LLM prompt tightened: explicit instruction against asterisks, backticks, underscores, and all Markdown.
v1.1.2
- `deliver.sh` Telegram payload no longer includes `parse_mode: Markdown` — LLM-composed text containing backticks or asterisks was causing Telegram to return a 400 parse error. Digests now deliver as plain text.
- `digest-agent.sh` prompt updated to request plain text output from the agent instead of Telegram Markdown, preventing the agent from generating formatting characters that would trigger Telegram parse errors.
v1.1.1
- Adds explicit mention of new log file: `last-delivered.md`, containing the actual digests sent.
- Clarifies separation between raw logs (`last-digest.md`) and delivered summaries (`last-delivered.md`).
- Highlights a privacy warning: raw logs are sent in full to OpenClaw agents, and may be transmitted off-host if a cloud LLM is used.
- Refines log handling explanations—details when and how files are written, and by which components.
- No functional/code changes; documentation improved for privacy, transparency, and log review processes.
v1.1.0
**Now uses OpenClaw agent to compose human-readable digests, with fallback.**
- Digest agent now calls your configured OpenClaw agent to generate summaries; if the agent call fails, falls back to a mechanical summary format.
- Full raw log for each digest run is saved as `~/.pulse-board/logs/last-digest.md` for review or agent access.
- Updated skill requirements: the `openclaw` CLI must be in your PATH for LLM-enhanced digests (old format used if absent).
- SKILL.md and descriptions updated for improved clarity and current behavior.
v1.0.5
**Expanded documentation to clarify all system changes and operational transparency:**
- Added a new "What this skill touches" section detailing exactly what Pulse Board modifies (filesystem, crontab, environment, and network) and all actions are explicitly confirmed.
- Clearly explains secrets handling: no silent writes, credentials never logged, and outbound connections only for digest delivery.
- Improved install instructions for user awareness and transparency.
- Discovery mode description clarified for interactive skill plug-in.
- No code or functional changes; documentation only.
v1.0.4
Initial release of pulse-board: a universal operational digest for agent skill stacks.
- Captures log lines from every scheduled skill via log-append.sh.
- Twice daily, digest-agent.sh summarizes all activity and delivers via Telegram, Discord, or local log.
- install.sh provides a one-command, interactive setup for directories, configuration, and secrets.
- plug.sh and unplug.sh offer simple registration/removal of skills and their cron jobs, including an interactive discovery mode.
- No root access or system path modification required; all config/data contained per user.
元数据
常见问题
Pulse Board 是什么？
Universal operational digest for agent skill stacks. Every scheduled skill logs its outcome with log-append.sh. Twice daily, digest-agent.sh reads the log, c... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 405 次。
如何安装 Pulse Board？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install pulse-board」即可一键安装，无需额外配置。
Pulse Board 是免费的吗？
是的，Pulse Board 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Pulse Board 支持哪些平台？
Pulse Board 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 Pulse Board？
由 LittleJakub（@littlejakub）开发并维护，当前版本 v1.1.3。
推荐 Skills