功能描述

Provide new contributors and agents with a concise tour of the workspace identity files (SOUL.md, USER.md, AGENTS.md, TOOLS.md) plus onboarding tips. Use when a newcomer needs context or when you want to double-check how this workspace is configured.

使用说明 (SKILL.md)

Context Onboarding

Name: Context Onboarding
Author: crimsondevil333333

When to use this skill

You're guiding someone new through Clawdy/Clawd and want a quick summary of the personality, operating rules, and per-skill notes.
You need to remind yourself of the tone preferences or tooling constraints without reading every document top to bottom.

What it does

scripts/context_onboarding.py reads the key documents (SOUL.md, USER.md, AGENTS.md, TOOLS.md by default) and prints the first few lines of each so you can skim character, rules, and tooling notes.
The CLI supports --files to include additional documents, --lines to control how many lines are shown per file, and --brief to emit only the opening sentence of each section.
Use references/context-guidelines.md when you need more guidance about what newcomers should read next or how to keep the vibe consistent.

CLI usage

python3 skills/context-onboarding/scripts/context_onboarding.py summarizes the default identity docs and prints the first five lines per file.
Add --files docs/PLAYBOOK.md docs/ROLE.md to weave in extra reference material that onboards the newcomer to cadence notes or release rituals.
Pair --lines 2 with --brief to emit single-line headlines when you just need the gist.
--workspace /path/to/other-workspace lets you compare multiple workspaces or prepare summaries for a sister repo before pairing.

Example command

python3 skills/context-onboarding/scripts/context_onboarding.py --files references/context-guidelines.md HEARTBEAT.md --lines 2

This prints the opening two lines for the personality files plus the heartbeat and onboarding guide so you can review vibe, reminders, and cadence expectations without opening every file.

Options

--files \x3Cpath>: Accepts extra markdown files (comma/space separated) that the script should include in the output order you provide.
--lines \x3Cn>: Controls how many lines from each file are shown (default 5) so you can tighten or loosen the briefing.
--brief: Shrinks each preview to the first sentence (splitting on ., ?, or !). Use this for lightning summaries during sync calls.
--workspace \x3Cdir>: Point the CLI at another workspace root; useful for onboarding clones, reviewing experimental docs, or prepping a new repo.

References

references/context-guidelines.md documents onboarding topics, role expectations, cadence notes, and reminders for how this group runs.

Resources

GitHub: https://github.com/CrimsonDevil333333/context-onboarding
ClawHub: https://www.clawhub.ai/skills/context-onboarding

安全使用建议

This is a small, coherent tool for summarizing workspace identity docs. Before running it: (1) note the minor documentation path mismatch (the script is at scripts/context_onboarding.py); (2) avoid pointing --workspace or --files at directories containing secrets or private keys — the tool will read whatever files you give it and print their contents; (3) review the short script if you want assurance it doesn't transmit data (it doesn't); and (4) run tests in a sandboxed environment if you are uncertain.

功能分析

Type: OpenClaw Skill Name: Developer: Version: Description: OpenClaw Agent Skill Suspicious High-Entropy/Eval files: 2 The `scripts/context_onboarding.py` skill is classified as suspicious due to its overly broad file access capabilities. While its stated purpose is to summarize workspace identity documents, the `--workspace` and `--files` arguments allow it to read arbitrary files from any directory accessible to the agent. This includes potentially sensitive files like user credentials or system configuration files (e.g., `~/.ssh/id_rsa`, `/etc/passwd`). Although the skill itself does not contain explicit malicious intent like data exfiltration or remote execution, it provides a high-risk primitive that could be leveraged by a malicious prompt injection against the OpenClaw agent to disclose sensitive information.

能力评估

✓ Purpose & Capability

Name/description match the delivered files: a small script and tests that summarize SOUL.md, USER.md, AGENTS.md, TOOLS.md. No unrelated credentials, binaries, or installs are requested.

ℹ Instruction Scope

SKILL.md instructs the agent to run a local Python CLI that reads user-specified files and a workspace path. That behavior is consistent with onboarding. Note: some documentation examples reference a 'skills/context-onboarding/...' path while the shipped script lives at 'scripts/context_onboarding.py' — a documentation/path inconsistency (not a security problem). Also, because the CLI accepts arbitrary --workspace and --files, it will read any files you point it to; this is expected but means you should avoid pointing it at directories with secrets.

✓ Install Mechanism

No install spec is provided (instruction-only plus small Python script). Nothing is downloaded or written to disk by an installer.

✓ Credentials

No environment variables, credentials, or config paths are requested. The script only reads files you explicitly pass or the default identity docs.

✓ Persistence & Privilege

The skill does not request permanent presence (always:false) and contains no code that modifies agent/system configs. Autonomous invocation is allowed by default but not combined with any other elevated privileges.

版本历史

v1.0.1

- Added a detailed "CLI usage" section describing how to use the script with various flags. - Included practical command examples to clarify real-world use of the CLI. - Expanded documentation of available CLI options, including `--files`, `--lines`, `--brief`, and `--workspace`. - Improved onboarding guidance to help users efficiently summarize workspace identity files and configuration.

v1.0.0

- Initial release of the context-onboarding skill. - Provides a concise tour of key workspace identity files (SOUL.md, USER.md, AGENTS.md, TOOLS.md) for new contributors and agents. - Adds a script to display the first few lines of each important file, helping users quickly review character, rules, and tooling information. - Supports CLI options to customize included files, number of lines shown, and brief output format. - Includes references for further onboarding guidance and workspace operation tips.

元数据

Slug context-onboarding

版本 1.0.1

许可证 —

累计安装 3

当前安装数 2

历史版本数 2

常见问题