Claude Code Usage

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.

ConcernHigh Confidence

ASI07: Insecure Inter-Agent Communication

What this means

If this helper is run, your Claude usage/reset notification could be sent to someone else's Telegram account.

Why it was flagged

This sends reset notification content to a fixed Telegram target ID instead of a user-provided or documented recipient.

Skill content

printf '%s' "$MESSAGE" | clawdbot message send --telegram --target 5259918241

Recommendation

Do not run monitor-and-notify.sh unless the hard-coded target is removed or replaced with your own explicit, documented configuration.

NoteHigh Confidence

ASI03: Identity and Privilege Abuse

What this means

Running the checker gives the script access to your local Claude Code OAuth token, although the shown code uses it for the stated Anthropic usage endpoint.

Why it was flagged

The script reads local Claude Code OAuth credentials and uses the access token to query Anthropic's usage API.

Skill content

security find-generic-password -s "Claude Code-credentials" -w ...; secret-tool lookup application "Claude Code" ...; -H "Authorization: Bearer $TOKEN"

Recommendation

Only run this from a trusted local copy, and verify that token handling remains limited to api.anthropic.com.

NoteHigh Confidence

ASI10: Rogue Agents

What this means

The skill may continue checking usage and sending reminders until you remove the scheduled job.

Why it was flagged

The monitoring feature intentionally creates persistent scheduled reminders that continue after initial setup.

Skill content

This creates a self-scheduling chain of cron jobs ... Repeats every 5 hours automatically

Recommendation

Enable monitoring only if you want recurring background checks, and know how to list and remove the related Clawdbot cron jobs.

NoteHigh Confidence

ASI05: Unexpected Code Execution

What this means

A usage check can automatically make a small Claude CLI request when credentials need refreshing.

Why it was flagged

When the stored token is expired, the script invokes the local claude CLI to trigger token refresh.

Skill content

echo "2+2" | claude >/dev/null 2>&1 || true

Recommendation

If you do not want automatic CLI refresh behavior, refresh Claude manually or remove this fallback before running the script.