WhatsApp FAQ Bot

AdvisoryAudited by Static analysis on Apr 30, 2026.

Overview

No suspicious patterns detected.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Users have less context about where the skill came from, so they should review the included script before installing.

Why it was flagged

The skill has limited provenance metadata, although the included artifacts are simple and no external dependency or install-time execution is shown.

Skill content
Source: unknown; Homepage: none; No install spec — this is an instruction-only skill.
Recommendation

Prefer a skill with a clear repository/homepage when available, or manually review the bundled Python file before use.

What this means

Incorrect or untrusted FAQ entries could later influence answers given to users or customers.

Why it was flagged

The FAQ knowledge base is persistent local state that can be reused for future answers; this is expected for a FAQ bot but users should control what content is imported.

Skill content
Data is stored in `~/.faq-bot/` by default. Override with `FAQ_BOT_DIR` environment variable.
Recommendation

Import only trusted FAQ files, periodically review the stored knowledge base, and set `FAQ_BOT_DIR` to an appropriate location if needed.

NoteHigh Confidence
ASI10: Rogue Agents
What this means

If wired into an auto-reply workflow, the bot may send FAQ-based responses automatically, including low-confidence or outdated answers.

Why it was flagged

The documentation explicitly supports autonomous or scheduled use for replies; this is purpose-aligned, but it can continue producing answers without per-message review if the user configures it that way.

Skill content
the agent can use the `search` command to find the best match and respond automatically. Configure it in your cron or agent system prompt.
Recommendation

Test search thresholds and fallback behavior before enabling automatic customer replies, and keep a manual review path for uncertain matches.