Back to skill

Security audit

Auto Updater 1

Security checks for vulnerabilities and agentic risk

Overview

This skill openly sets up automatic updates, but it should be reviewed because it creates a persistent daily job that can change Clawdbot and every installed skill without fresh approval each time.

Install only if you intentionally want unattended daily updates to Clawdbot and every installed skill. Prefer running a dry run first, limiting updates to trusted skills or approved versions, using a least-privileged account, and confirming you know how to inspect and remove the cron job.

Vulnerability Patterns
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Findings (3)

T06 · System Persistence

Error
Location
references/agent-guide.md:76
Finding

Persistent Daily Update Task Executes Across Sessions

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Error
Location
references/agent-guide.md:41
Finding

Unattended Installation of Mutable Latest Packages and Skill Updates

Content
View full analysis
/dev/null && npm list -g clawdbot &> /dev/null; then npm update -g clawdbot@latest 2>&1 | tee -a "$LOG_FILE" elif command -v pnpm &> /dev/null && pnpm list -g clawdbot &> /dev/null; then pnpm update -g clawdbot@latest 2>&1 | tee -a "$LOG_FILE" elif command -v bun &> /dev/null; then bun update -g clawdbot@latest 2>&1 | tee -a "$LOG_FILE" else log "Running clawdbot update (source install)" clawdbot update 2>&1 | tee -a "$LOG_FILE" || true fi # Run doctor for migrations log "Running doctor..." clawdbot doctor --yes 2>&1 | tee -a "$LOG_FILE" || true # Capture new version CLAWDBOT_VERSION_AFTER=$(clawdbot --version 2>/dev/null || echo "unknown") # Update skills log "Updating skills via ClawdHub..." SKILL_OUTPUT=$(clawdhub update --all 2>&1) || true echo "$SKILL_OUTPUT" >> "$LOG_FILE" ``` ### Technical Analysis The `clawdbot@latest` reference is mutable and can resolve to a release that did not exist when the Skill was audited. Similarly, `clawdhub update --all` accepts future releases for every installed Skill. The workflow does not pin approved versions, verify cryptographic hashes or signatures, constrain trusted publishers, inspect package lifecycle behavior, present release diffs, or require approval before installation. Running `clawdbot doctor --yes` also permits migrations to proceed without interactive review. This creates a supply-chain exposure: the effective code and Skill instructions executed in future runs can change after the current package has been reviewed. ### Attack Path 1. An attacker compromises a ...[truncated 1241 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
references/agent-guide.md:54
Finding

Critical Update and Migration Failures Are Suppressed

Content
View full analysis
&1 | tee -a "$LOG_FILE" || true # Capture new version CLAWDBOT_VERSION_AFTER=$(clawdbot --version 2>/dev/null || echo "unknown") # Update skills log "Updating skills via ClawdHub..." SKILL_OUTPUT=$(clawdhub update --all 2>&1) || true echo "$SKILL_OUTPUT" >> "$LOG_FILE" log "Auto-update complete." ``` ### Technical Analysis The script begins with `set -e`, which would normally terminate execution after an unhandled command failure. Appending `|| true` explicitly overrides that protection for `clawdbot doctor --yes` and `clawdhub update --all`. As a result, failed migrations and failed Skill updates are converted into successful shell control flow. The unconditional `log "Auto-update complete."` entry does not distinguish complete success from partial or total failure. Reliable detection is left to later parsing of free-form command output rather than preserved exit codes and structured status. An attacker controlling update output could also make a failure less obvious by emitting misleading text, although the audited files do not contain evidence that such an attacker is currently present. ### Attack Path 1. A Clawdbot migration or Skill update fails because of corruption, incompatibility, network interruption, permission denial, or malicious update behavior. 2. `|| true` discards the nonzero exit status. 3. The script continues despite the failed security-sensitive operation. 4. The script writes `Auto-update complete` and emits its normal summary markers. 5. A reporting Agent or operator may interpret the ru ...[truncated 691 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Self-Modification

High
Category
Rogue Agent
Confidence
98% confidence
Finding

The skill explicitly performs clawdhub update --all, which is direct self-modification of the agent's installed skill set. This is dangerous because it grants scheduled, unattended authority to replace or alter agent capabilities, creating a strong supply-chain compromise path and increasing the blast radius of a malicious or broken skill update.

Content

Scanner excerpt · references/agent-guide.md (reported line 61)May include surrounding context.

md
# Capture new version
CLAWDBOT_VERSION_AFTER=$(clawdbot --version 2>/dev/null || echo "unknown")

# Update skills
log "Updating skills via ClawdHub..."
SKILL_OUTPUT=$(clawdhub update --all 2>&1) || true
echo "$SKILL_OUTPUT" >> "$LOG_FILE"

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill sets up unattended recurring updates that will modify the Clawdbot installation and all installed skills, but the user-facing description does not prominently warn that software and skill files will be changed automatically on a schedule. This can lead to users enabling persistent self-modification without fully understanding the trust and supply-chain implications, especially because clawdhub update --all pulls and applies remote changes across all skills.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The guide instructs the agent to perform unattended updates to the core bot and all installed skills, which can change code, behavior, dependencies, and configuration without any explicit warning, approval gate, or rollback guidance. In an agent context, automatic self-updating materially increases supply-chain and operational risk because compromised packages or breaking changes can be applied silently on a schedule.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
81% confidence
Finding

The guide recommends creating a persistent helper script under ~/.clawdbot/scripts/auto-update.sh, which establishes durable executable state that can be reused by future runs. In combination with cron-based execution and update authority, this persistence increases risk because any later tampering with that script can repeatedly execute under the user's automation flow.

Content

Scanner excerpt · references/agent-guide.md (reported line 21)May include surrounding context.

bun pm ls -g 2>/dev/null | grep clawdbot && echo "bun-global"

text

## Step 2: Create the Update Script (Optional)

For complex setups, create a helper script at `~/.clawdbot/scripts/auto-update.sh`:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The customization example allows delivery of update summaries to external providers such as Telegram, but it does not warn that logs or update results may contain sensitive operational details, package names, errors, paths, or other metadata. This can unintentionally disclose environment information to third-party services or recipients if misconfigured.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The setup confirmation example states that daily updates will run at 4:00 AM in America/Los_Angeles. This natural-language example enforces a specific locale/time zone and does not indicate user opt-in, configurability, or a region-specific requirement.

Content

No source excerpt is available for this finding.