Back to skill

Security audit

ulta-research

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its Ulta research purpose, but its helper can send undocumented POST requests with the user's Crawlora API key.

Install only if you are comfortable giving this helper access to your Crawlora API key and sending Ulta-related queries to Crawlora. Prefer a revised version that rejects POST, validates product IDs as a single path segment, and exposes only the documented GET endpoints before using it in automated or untrusted-content workflows.

Vulnerability Patterns
  • 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
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/crawlora.sh:44
Finding
Helper Permits Undocumented POST Requests and Overly Broad Product Routes<![CDATA[ ## Vulnerability Details **File Location**: `scripts/crawlora.sh`, lines 44–68 and 99–104 **Vulnerability Type**: Excessive API capabilities and insufficient route validation **Risk Level**: Medium ### Vulnerable Code ```bash case "$method" in GET|POST) ;; *) echo "only GET and POST are supported by the ulta-research skill" >&2 exit 2 ;; esac # Reject path syntax that could smuggle a route through a shell glob check. case "$path" in ""|*[?#%]*|*..*|*//* ) echo "invalid path for the ulta-research skill" >&2 exit 2 ;; esac case "$path" in /ulta/categories) ;; /ulta/category) ;; /ulta/product/*) ;; /ulta/product/questions) ;; /ulta/product/reviews) ;; /ulta/search) ;; /ulta/stores) ;; /ulta/suggest) ;; *) echo "path is not in the ulta-research skill catalog" >&2 exit 2 ;; esac ``` ```bash else [ -n "$body" ] || body="${rest[0]:-}" [ -n "$body" ] || body='{}' # Stream the body on stdin so curl never interprets a user value as its # @file shorthand (and cannot read local files supplied in a request body). printf '%s' "$body" | curl -fsS -X "$method" "${auth[@]}" \ -H "Content-Type: application/json" --data-binary @- "${base}${path}" fi ``` ### Technical Analysis The endpoint reference defines all eight operations exposed by this Skill as `GET` requests. Nevertheless, the helper accepts both `GET` and `POST` for every allowlisted path and forwards an arbitrary caller-controlled JSON body for POST requests. The route allowlist is also broader than the documented interface. The pattern `/ulta/product/*` accepts any descendant beginning with `/ulta/product/`, rather than accepting only a single validated product identifier. For example, additional undocumented path segments can pass the local check. Together, these behaviors violate least-capability principles: the helper exposes HTTP methods and route shapes that are not required for the declared Ulta research functionality. Requ ...[truncated 2155 chars]
Remediation
<![CDATA[ ## Remediation Suggestions 1. Restrict the helper to the only documented method: ```bash if [ "$method" != "GET" ]; then echo "only GET is supported by the ulta-research skill" >&2 exit 2 fi ``` 2. Remove POST body parsing and the POST-specific curl branch entirely. Eliminating unused functionality is safer than retaining dormant generic request support. 3. Validate product-detail routes as exactly one product identifier. For example: ```bash case "$path" in /ulta/categories|\ /ulta/category|\ /ulta/product/questions|\ /ulta/product/reviews|\ /ulta/search|\ /ulta/stores|\ /ulta/suggest) ;; /ulta/product/[A-Za-z0-9_-]*) product_id="${path#/ulta/product/}" case "$product_id" in ""|*[!A-Za-z0-9_-]*) echo "invalid Ulta product ID" >&2 exit 2 ;; esac ;; *) echo "path is not in the ulta-research skill catalog" >&2 exit 2 ;; esac ``` 4. Prefer separate wrapper functions for each endpoint rather than accepting a generic method and path. Each wrapper should allow only documented parameters and validate numeric fields, pagination, coordinates, radius, SKU, and product IDs. 5. Add regression tests confirming that POST, nested product descendants, encoded path separators, query fragments embedded in paths, and unsupported routes are rejected before curl executes. 6. Retain the existing fixed HTTPS base URL, API-key character validation, private temporary configuration, cleanup trap, and rejection of curl `@file` query syntax. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (8)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill instructs use of a shell helper script (`scripts/crawlora.sh`) but does not declare any tool restrictions or allowed-tool scope. In an agent environment, this creates unnecessary ambiguity about what execution capabilities the skill may use, increasing the chance of overbroad shell access or unsafe command invocation beyond the intended API calls.

External Transmission

Medium
Category
Data Exfiltration
Content
#!/usr/bin/env bash
# Crawlora REST helper — minimal, dependency-free (curl only).
# Calls https://api.crawlora.net/api/v1 with your Crawlora API key.
# Get a free key (2,000 credits/mo, no card) at https://crawlora.net?utm_source=github&utm_medium=referral&utm_campaign=crawlora-skills.
#
Confidence
70% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
#!/usr/bin/env bash
# Crawlora REST helper — minimal, dependency-free (curl only).
# Calls https://api.crawlora.net/api/v1 with your Crawlora API key.
# Get a free key (2,000 credits/mo, no card) at https://crawlora.net?utm_source=github&utm_medium=referral&utm_campaign=crawlora-skills.
#
# Usage:
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
#!/usr/bin/env bash
# Crawlora REST helper — minimal, dependency-free (curl only).
# Calls https://api.crawlora.net/api/v1 with your Crawlora API key.
# Get a free key (2,000 credits/mo, no card) at https://crawlora.net?utm_source=github&utm_medium=referral&utm_campaign=crawlora-skills.
#
# Usage:
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
#!/usr/bin/env bash
# Crawlora REST helper — minimal, dependency-free (curl only).
# Calls https://api.crawlora.net/api/v1 with your Crawlora API key.
# Get a free key (2,000 credits/mo, no card) at https://crawlora.net?utm_source=github&utm_medium=referral&utm_campaign=crawlora-skills.
#
# Usage:
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
#!/usr/bin/env bash
# Crawlora REST helper — minimal, dependency-free (curl only).
# Calls https://api.crawlora.net/api/v1 with your Crawlora API key.
# Get a free key (2,000 credits/mo, no card) at https://crawlora.net?utm_source=github&utm_medium=referral&utm_campaign=crawlora-skills.
#
# Usage:
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Content
# Keep the API key out of the curl process command line. A private temporary
# config supplies the header and is removed automatically on exit.
curl_config="$(mktemp "${TMPDIR:-/tmp}/crawlora-curl.XXXXXX")"
chmod 600 "$curl_config"
trap 'rm -f "$curl_config"' EXIT
printf 'header = "x-api-key: %s"\n' "$CRAWLORA_API_KEY" >"$curl_config"
auth=(--config "$curl_config")
Confidence
80% confidence
Finding
Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

External Transmission

Medium
Category
Data Exfiltration
Content
[ -n "$body" ] || body='{}'
  # Stream the body on stdin so curl never interprets a user value as its
  # @file shorthand (and cannot read local files supplied in a request body).
  printf '%s' "$body" | curl -fsS -X "$method" "${auth[@]}" \
    -H "Content-Type: application/json" --data-binary @- "${base}${path}"
fi
Confidence
70% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Static analysis

No suspicious patterns detected.