Back to skill

Security audit

garmin-nutrition

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Garmin food-log helper that stores nutrition data locally and can sync or delete Garmin entries only through explicit commands.

Install only if you are comfortable letting the skill read your existing Garmin OAuth session and, when you approve --yes without --no-garmin, upload meal names, calories, macros, meal slots, and timestamps to Garmin. Use --no-garmin for a strictly local journal and review dry-run output before deletes or corrections.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (7)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 62)May include surrounding context.

md
| Read day | `GET /nutrition-service/food/logs/{date}` (via `get_nutrition_daily_food_log`) |
| Settings | `GET /nutrition-service/settings/{date}` |
| Add | `PUT /nutrition-service/food/logs/quickAdd`, body `{mealDate, quickAddItems:[…]}` |
| Delete | `DELETE /nutrition-service/food/logs/{date}`, body `{"logIds": [id]}` → 204 |

Two things worth knowing if you build on this:

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · SKILL.md (reported line 130)May include surrounding context.

md
## Garmin API notes

Private, undocumented endpoints; verified August 2026. Add = `PUT /nutrition-service/food/logs/quickAdd`. Delete = `DELETE /nutrition-service/food/logs/{date}` with `{"logIds": [...]}` (date in path, ids in body — an id in the path 404s). `quickAdd` with `action: "DELETE"` returns 200 and does nothing; never use it. There is no update. After every write the script re-reads the log to confirm; `unknown_after_push` in the output means check before retrying, a retry can duplicate.

## Related

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/garmin_nutrition.py (reported line 13)May include surrounding context.

python
an interrogation. The journal is the source of truth; Garmin is a sink.

Garmin food logging is a Connect+ feature. The write path uses Garmin's
private Quick Add endpoint; deletion uses DELETE /food/logs/{date} with
{"logIds": [...]} — the quickAdd action=DELETE form returns 200 and does
nothing, so it is never used here.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill explicitly describes capabilities to read environment variables, read/write local files, read OAuth tokens from ~/.garminconnect/, and make network calls, but it does not declare any explicit tool scope or permissions. That mismatch weakens user and platform oversight because the effective authority of the skill is broader than what a reviewer can verify from metadata.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The guidance and example question explicitly prescribe replying in Russian, including a fixed Russian disambiguation prompt. This creates a locale/language constraint in the skill instructions without stating that the user chose Russian or that the skill is region-specific.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The user-facing help strings include specific Russian examples such as 'пачка', 'масло', 'помидоры', and 'фета'. This can impose a locale expectation in natural-language interaction without explicit user opt-in or a documented language choice.

Content

No source excerpt is available for this finding.