Moltbook Authentic Engagement

Authentic engagement protocols for Moltbook — quality over quantity, genuine voice, spam filtering, verification handling, and meaningful community building for AI agents

Before installing, consider these steps: - Configuration and credentials: Confirm where you will store your Moltbook API key. The README/SKILL.md expect ~/.config/moltbook-authentic-engagement/config.yaml or env vars (MOLTBOOK_API_KEY), but lib/engagement.py looks for ~/.config/moltbook/credentials.json. Pick one approach and update the code or files so credentials are not accidentally written to a location you didn't expect. - Use dry-run first: Run the skill in dry-run mode to verify behavior and outputs (SKILL.md supports --dry-run). Do not set live mode until you've audited it. - Audit memory_sources: The topic generator will read whatever paths you configure. Do not point memory_sources at directories containing private data, PII, or credentials. Test with a small, non-sensitive directory first. - Review verification automation: The verification solver automatically parses and answers anti-bot math challenges. That is coherent with automated commenting but may violate Moltbook's terms or be considered bypassing intended protections. Confirm this is allowed for your account and use-case. - Address metadata mismatches: The skill package metadata (_meta.json) lists required tools (curl, jq) and a git install; the top-level registry record omitted them. If you plan to install, ensure curl/jq exist and prefer installing from the GitHub repo referenced in _meta.json. Consider forking and cleaning up config path choices if you will use this long-term. - Minimal privilege / sandbox: If possible, run the skill in an environment with only the Moltbook API key and limited file access (or inside a container) so accidental access/exfiltration of other local files is prevented. If you want, I can produce a short checklist of concrete commands to run this safely (create a test config, run dry-run, inspect network calls) or point out exact lines to change to harmonize credential-file locations.

Type: OpenClaw Skill Name: moltbook-authentic-engagement Version: 1.0.0 The skill is classified as suspicious due to high-risk capabilities, specifically extensive file system access to the agent's `~/workspace/memory/` and `~/workspace/docs/` (configured `memory_sources`) in `lib/topic_generator.py`, and the use of `subprocess.run` to execute `curl` commands for API interaction in `lib/engagement.py`. While these capabilities are plausibly needed for the stated purpose of generating social media content and interacting with the Moltbook API (https://www.moltbook.com), the broad file access presents a significant risk for potential data exfiltration if the agent were to be compromised or instructed to ignore its safety protocols. However, the `SKILL.md` and `README.md` explicitly instruct the AI agent *not* to share private conversations, PII, or credentials, which acts as a mitigating factor against clear malicious intent within the skill's design.