Slack Channel Context

Automatically loads Slack channel context files (e.g., bebops.md, C0AK8SDFS4W.md) into session context for Slack channels and threads. Use this skill wheneve...

This skill is internally consistent and appears safe to install, but review the following before enabling: (1) it will read any Markdown files in ~/.openclaw/workspace/slack-channel-contexts/ (or a custom SLACK_CONTEXT_CONTEXTS_DIR) and inject that text into agent session context — do not store secrets or sensitive data in those files; (2) the skill may create the contexts directory and write a README.md there the first time it runs; (3) the skill is a helper library and does not itself connect to Slack — your Slack integration must call load_channel_context and pass message metadata; (4) documentation has minor inconsistencies (cache TTL default and .env handling), so verify environment variables are set to the values you expect. If you want to be cautious, inspect the skill's scripts (present in the package) and disable it in openclaw.json or set SLACK_CONTEXT_ENABLED=false until you're ready.

Type: OpenClaw Skill Name: slack-channel-context Version: 1.0.0 The skill is designed to automatically load Slack channel-specific context from local Markdown files into an AI session. It is classified as suspicious because the implementation in `scripts/skill.py` and `scripts/load_context.py` lacks input sanitization on the `channel_id` and `channel_name` parameters, which are used to construct file paths. This creates a path traversal vulnerability that could allow an attacker to read arbitrary `.md` files on the host system if the agent is provided with a malicious channel identifier (e.g., `../../etc/shadow`). No evidence of intentional malice, data exfiltration, or unauthorized remote control was found.