Back to skill

Security audit

librelinkup-monitor

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because scanners report safety and credential-handling weaknesses in its glucose-monitoring workflow; one image was not fully inspected, so evidence is incomplete.

Consider the reported stale-reading risk and how credentials and glucose data are handled before installing. The scanners did not fully inspect one referenced image, so the results do not establish that every part of the skill is safe.

Vulnerability Patterns
  • 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
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Rogue AgentSelf-Modification, Session Persistence
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (8)

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 9)May include surrounding context.

md
icon: "assets/icon.png"

Stale readings are treated as successful polls despite the documented stale-sensor failure handling.

High
Category
Excessive Agency
Confidence
99% confidence
Finding

The poller extracts glucose value and trend without checking the measurement timestamp, then clears the failure history and evaluates the reading as current. A cloud endpoint returning an old normal reading can therefore keep emitting QUIET indefinitely while actual glucose becomes dangerously low, undermining the monitor's central safety function.

Content

No source excerpt is available for this finding.

The setup workflow requests an account password in chat without warning about credential exposure through conversation storage or processing.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The setup prompt explicitly asks the user to give an account password to the conversational agent, potentially exposing it through model processing, conversation history, and logging. Saving it locally without echoing it does not undo that exposure, and the caregiver account provides access to sensitive patient glucose information.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
88% confidence
Finding

Persistent configuration is legitimate for scheduled monitoring, but the instructions store a reusable account password in plaintext without specifying restrictive file permissions or access controls. Depending on local defaults and backup practices, other users or processes could obtain credentials that expose patient telemetry; this is insecure credential storage, not malicious session persistence.

Content

Scanner excerpt · SKILL.md (reported line 108)May include surrounding context.

text

### Step 3: Configure Credentials & Thresholds
Create `libre_creds.json` in your Hermes home directory (e.g. `~/.config/librelinkup/creds.json` or `%LOCALAPPDATA%\hermes\libre_creds.json`):

```json
{

The alert workflow sends sensitive glucose readings through messaging services without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The documented workflow routes glucose readings through external messaging providers and can escalate to another caregiver without an explicit health-data disclosure and recipient-consent step. Although external notifications are an intended feature rather than covert exfiltration, misconfigured recipients, provider retention, or notification previews can expose sensitive medical information.

Content

No source excerpt is available for this finding.

Tainted flow: 'failfile' from os.getenv (line 204, credential/environment) → open (file write)

Medium
Category
Data Flow
Confidence
65% confidence
Finding

Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.

Content

Scanner excerpt · scripts/glucose_watch.py (reported line 109)May include surrounding context.

python
except Exception:
        n = 1
    try:
        with open(failfile, "w") as fh:
            fh.write(str(n))
    except Exception:
        pass

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · FAQ.md (reported line 26)May include surrounding context.

md
### 3. ASSUMPTION OF RISK & INDEMNIFICATION (HOLD HARMLESS)
**BY INSTALLING, RUNNING, OR RELYING UPON THIS SOFTWARE, YOU EXPRESSLY ACKNOWLEDGE AND AGREE THAT YOU DO SO ENTIRELY AT YOUR OWN RISK.**

To the maximum extent permitted by applicable law, the authors, maintainers, contributors, and copyright holders of this repository (including Nous Research, Syzygy Studios, and individual developers) **DISCLAIM ALL WARRANTIES OF ANY KIND, WHETHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OPERATIONAL AVAILABILITY, ACCURACY, AND NON-INFRINGEMENT.**

**IN NO EVENT SHALL THE AUTHORS, MAINTAINERS, OR AFFILIATES BE LIABLE FOR ANY CLAIM, DAMAGES, LOSSES, INJURIES, HOSPITALIZATIONS, ADVERSE HEALTH EVENTS, OR LIABILITIES**, WHETHER IN AN ACTION OF CONTRACT, TORT (INCLUDING NEGLIGENCE), STRICT LIABILITY, OR OTHERWISE, ARISING FROM, OUT OF, OR IN CONNECTION WITH THE SOFTWARE, THE USE OF THE SOFTWARE, DELAYED NOTIFICATI