← 返回 Skills 市场

Cross-Channel & Cross-Agent Skill Sharing

Name: Cross-Channel & Cross-Agent Skill Sharing
Author: skglau

作者 Gabriel · GitHub ↗ · v0.1.1 · MIT-0

cross-platform ⚠ suspicious

105

总下载

当前安装

版本数

在 OpenClaw 中安装

/install cross-channel-cross-agent-sharing

功能描述

Ensure new capabilities gained from workspace-installed tools/packages are propagated across channels and agent sessions. Use when any tool/env/package insta...

使用说明 (SKILL.md)

Cross-Channel & Cross-Agent Skill Sharing

Execute this workflow whenever workspace capability changes

Detect capability change source:
- New package/tool installed in workspace
- Existing tool upgraded and behavior changed
- New env/interpreter path required for capability
Update TOOLS.md with operational facts:
- Exact runtime/interpreter path
- Install command
- Verify command
- Package versions
- One-line usage note
Create or update a reusable skill when pattern is repeatable:
- Add/adjust SKILL.md
- Add scripts under scripts/ for deterministic steps
- Keep instructions concise and invocation-focused
- Prefer adding a verification script for fast capability checks
Log event in memory/YYYY-MM-DD.md:
- What changed
- How to verify
- Any limits/known caveats
Inform current conversation:
- Summarize new capability
- Show exact invocation command or trigger phrase
Propagate to active related sessions/agents:
- If likely relevant, send brief sync via sessions_send
- Include only invocation + critical caveat

Required output format in chat (short)

New capability: \x3Cwhat is newly possible>
Use it with: \x3Cexact command / trigger>
Scope: Applies to all channels and sessions using this workspace
Caveat: \x3Cif any>

Quick Verification Helper

Use bundled script to verify interpreter + modules before claiming capability:

scripts/check_capability.sh \x3Cpython-bin> \x3Cmodule1> [module2 ...]

Example:

scripts/check_capability.sh ~/.openclaw/workspace/.venv-img/bin/python cv2 PIL numpy

Guardrails

Do not assume global python/node paths; prefer workspace-scoped env paths
Do not claim capability before running verify command
Keep cross-session notices minimal to avoid spam
If capability is channel-specific (provider restriction), state that explicitly

安全使用建议

This skill is coherent with its declared goal (propagate workspace capability changes), but it has a high blast radius because it tells the agent to edit shared docs/skills and notify other sessions. Before installing: (1) Confirm you trust any agent that can run this — consider requiring manual approval for applied changes or diffs to SKILL.md/scripts. (2) Ensure file-write permissions are scoped (e.g., restrict which directories can be modified). (3) Require human review of any proposed SKILL.md or script changes and retain backups/revision history for TOOLS.md and SKILL.md. (4) Limit or audit the use of session-wide notifications (sessions_send) to avoid unwanted propagation. If you cannot enforce these constraints, treat the skill as potentially risky.

功能分析

Type: OpenClaw Skill Name: cross-channel-cross-agent-sharing Version: 0.1.1 The skill bundle is a utility designed to help an AI agent manage and synchronize workspace capabilities, such as newly installed packages or tools, across different sessions. It contains a helper script (scripts/check_capability.sh) for verifying Python environments and instructions (SKILL.md) for updating workspace documentation and notifying other agent sessions. The logic is transparent, aligned with its stated purpose, and shows no signs of malicious intent, data exfiltration, or unauthorized persistence.

能力评估

ℹ Purpose & Capability

The name/description match the instructions: the skill is intended to detect workspace capability changes and propagate them by updating docs, creating/updating reusable skills, adding verification scripts, logging to memory, and notifying sessions. The requested resources (none: no env vars, no binaries, no install) are consistent with an instruction-only workflow that operates on workspace files. This is coherent with the declared purpose.

⚠ Instruction Scope

SKILL.md explicitly instructs the agent to modify workspace files (TOOLS.md, SKILL.md under reusable skills, scripts/), write memory/YYYY-MM-DD.md, and send cross-session notifications (sessions_send). Those actions go beyond simple observation/verification and give the agent permission to alter shared artifacts and notify other agents/sessions. The verification script itself is benign, but the overall instructions allow broad file writes and cross-session messaging — a scope that should be intentionally limited and audited.

✓ Install Mechanism

No install spec; this is instruction-only with a small local verification script. Nothing is downloaded or extracted from external URLs and no new binaries are installed by the skill, which is the lowest-risk install posture.

✓ Credentials

The skill requests no environment variables, no credentials, and no special config paths. It uses only workspace files and a local verification script that checks Python modules. The lack of requested secrets is appropriate for the stated functionality.

⚠ Persistence & Privilege

Although always:false (not force-included) and no credentials requested, the skill's workflow instructs creating/updating SKILL.md and scripts for reusable skills and propagating those changes across channels/sessions. This is a persistence/propagation capability: if misused, it can alter other skills' behavior or spread changes widely. Because autonomous invocation is allowed by default, consider whether this propagation behavior should require explicit human approval or extra gating.

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 cross-channel-cross-agent-sharing
安装完成后，直接呼叫该 Skill 的名称或使用 /cross-channel-cross-agent-sharing 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

v0.1.1

Add capability verification helper script and verification guidance

v0.1.0

Initial release: workspace capability propagation across channels and sessions

元数据

Slug cross-channel-cross-agent-sharing

版本 0.1.1

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 2

常见问题