Firm Security Audit

ReviewAudited by ClawScan on May 10, 2026.

Overview

This looks like a real security-audit skill, but it may automatically share audit results to Slack and urge immediate deployment changes without enough confirmation or redaction.

Before installing, verify the external MCP extension and Slack tool, use a least-privilege Slack app, require manual approval before sending audit results or applying fixes, and run remediation only after reviewing diffs and backups.

Findings (4)

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

Sensitive vulnerability details or configuration-derived information could be posted to a Slack channel automatically.

Why it was flagged

The skill checks local environment/config files and then instructs the agent to automatically send the audit results to Slack, with no redaction, recipient validation, or confirmation step described.

Skill content
"env_file_path": "/path/to/.env" ... "tool": "firm_export_slack_digest" ... "content": "<résultats de l'audit>", "channel": "#security-alerts"
Recommendation

Require explicit user approval before sending any report, redact secrets and file contents, limit the Slack channel and bot scope, and send only the minimum necessary summary by default.

What this means

An agent following the skill too literally could change gateway exposure, configuration files, or running containers in a way that disrupts service.

Why it was flagged

The remediation guidance encourages immediate network/deployment changes based on audit outputs, but does not define an explicit approval, backup, dry-run, or rollback process.

Skill content
Si CRITICAL (funnel actif sans rate limiter) : appliquer `fix_nginx` ou `fix_caddy` **immédiatement** ... Désactiver Funnel et appliquer Nginx/Caddy immédiatement ... Appliquer fix + redémarrer le container
Recommendation

Make remediation a separate, user-confirmed step; require a diff, backup, target path confirmation, and rollback instructions before changing deployment files or service state.

What this means

The agent may post messages using a Slack bot or user identity, potentially to a team-visible channel.

Why it was flagged

Posting security alerts to Slack is purpose-aligned, but it implies delegated Slack workspace authority that users should verify before enabling.

Skill content
firm_export_slack_digest       — notification équipe si CRITICAL trouvé
Recommendation

Use a dedicated low-privilege Slack app, restrict it to the intended channel, and document the credential/source of that authority.

What this means

The safety of the skill depends on tools that are not included in the reviewed artifact.

Why it was flagged

The skill depends on an external MCP extension and tool implementations, but the supplied package has no install spec or code to review.

Skill content
requires:
      - mcp-openclaw-extensions >= 2.0.0
Recommendation

Install the extension only from a trusted source, pin an exact reviewed version, and review the MCP tool behavior before granting it deployment or Slack access.