Signal

Security checks across malware telemetry and agentic risk

Overview

This is mostly a disclosed Signal messaging integration, but it documents Signal-triggered configuration changes enabled by default, which users should review before installing.

Before installing, use a dedicated Signal bot number, restrict `dmPolicy`, `allowFrom`, `groupPolicy`, and `groupAllowFrom`, consider setting `configWrites` to false, keep history and attachment handling minimal, and install `signal-cli` only from a trusted source.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

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.

#
ASI02: Tool Misuse and Exploitation
Medium
What this means

A paired or allowed Signal user may be able to change agent/channel configuration through chat commands, affecting future behavior beyond a single message.

Why it was flagged

This indicates Signal-origin commands can change OpenClaw configuration by default. The visible artifact does not clearly bound this to the owner, require explicit approval, or describe containment/rollback.

Skill content
- `configWrites`: Allow Signal channel to accept `/config set|unset` (default true).
Recommendation

Disable `configWrites` unless needed, restrict it to owner-only DMs, and require explicit confirmation and logging for every configuration change.

#
ASI03: Identity and Privilege Abuse
Medium
What this means

Messages and reactions may be sent from the configured Signal account, so mistakes or misuse can affect real conversations.

Why it was flagged

The skill uses a Signal account identity to send and react to messages. That is expected for the integration and is disclosed, but it is delegated account authority.

Skill content
The gateway connects to a *Signal device* (the `signal-cli` account). If you use your personal Signal account, the bot will ignore your own messages... Use a **separate bot number** for optimal operation.
Recommendation

Use a dedicated bot number, avoid using a personal Signal account, and keep DM/group allowlists narrow.

#
ASI06: Memory and Context Poisoning
Medium
What this means

Recent group chat content may be exposed to the agent and could steer its responses, including content from non-owner participants.

Why it was flagged

The skill documents including recent Signal group messages as agent context. This is purpose-aligned, but private or untrusted chat content may influence the agent.

Skill content
- Group history context uses `channels.signal.historyLimit` (default 50, set 0 to disable).
Recommendation

Set conservative history limits, disable history where not needed, and keep group access allowlisted.

#
ASI10: Rogue Agents
Low
What this means

Signal connectivity may rely on a background process that continues to receive or handle messages while the channel is enabled.

Why it was flagged

The skill documents automatic background daemon startup. It is disclosed and related to Signal connectivity, but users should know it may keep running outside a single interaction.

Skill content
- `autoStart`: Auto-spawn daemon (default true if `httpUrl` unset).
Recommendation

Disable `autoStart` if you want to manage the daemon manually, or configure a known `httpUrl` and monitor the daemon lifecycle.

#
ASI04: Agentic Supply Chain Vulnerabilities
Low
What this means

Users must obtain and configure `signal-cli` themselves, so safety depends on installing it from a trusted source.

Why it was flagged

The skill depends on an external `signal-cli` binary, while the provided install metadata has no install spec or required binary declaration. This is purpose-aligned but leaves dependency provenance to the user.

Skill content
- `cliPath`: Path to `signal-cli` binary.
Recommendation

Install `signal-cli` only from trusted upstream sources, keep it updated, and verify the configured `cliPath` points to the intended binary.