Back to skill

Security audit

Bug Hunter

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent bug-reporting helper, but its API debugging guidance can capture and persist sensitive request and response data without redaction or cleanup guidance.

Review this skill before installing if you may use it with production APIs or customer data. If used, redact Authorization headers, cookies, tokens, API keys, PII, and confidential request or response bodies before saving or sharing bug reports, and avoid the fixed /tmp/api-debug.log pattern.

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
SKILL.md:172
Finding

Predictable Temporary File May Expose Sensitive HTTP Debug Data

Content
View full analysis
&1 | tee /tmp/api-debug.log # Key things to capture: # - Request URL, method, headers, body # - Response status code, headers, body # - Timing information # - Any redirect chain ``` ### Technical Analysis The skill recommends capturing verbose `curl` output, request and response headers, and request and response bodies in the fixed path `/tmp/api-debug.log`. In realistic debugging sessions, this material can contain authorization headers, session cookies, API tokens, personal data, or other confidential application content. A predictable path in a shared temporary directory creates several risks: - Another local user may monitor or read the resulting file if its permissions are insufficiently restrictive. - An attacker may pre-create `/tmp/api-debug.log` as a symbolic link, potentially causing `tee` to overwrite a file writable by the process running the command. - A previous capture may be overwritten or remain accessible after debugging. - The use of `curl -v` increases the amount of potentially sensitive protocol metadata recorded. - The instructions provide no redaction, secure file creation, permission control, retention limit, or cleanup requirement. The issue does not independently grant elevated privileges. Exploitation is constrained by the operating-system permissions of the user executing the command and by the attacker's local access. ### Attack Path 1. A local attacker predicts that the documented debugging workflow will write to `/tmp/api-debug.log`. 2. The attacker monitors that path or pre-creates it as a file or symbolic link. 3. An operator or a ...[truncated 1188 chars]
Remediation
View remediation
"$debug_file" printf 'Diagnostic output saved to: %s\n' "$debug_file" ``` This safer example should still be accompanied by explicit instructions to omit secrets, sanitize captured content, and securely delete the file after use. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (4)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly tells users to include the full request and response in bug reports, but gives no warning to redact secrets or personal data. In practice, API headers, cookies, bearer tokens, session IDs, request bodies, and response payloads often contain sensitive information that could be copied into tickets, chats, or logs and then exposed to unintended audiences.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The example command uses verbose curl output and saves the complete request/response exchange to a local file via tee, which can persist credentials and sensitive business data on disk. Even if intended for debugging, this creates a reusable artifact that may later be read by other users, uploaded, or committed accidentally.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 177)May include surrounding context.

bash
# Save full request + response for bug report
curl -v -X POST "https://api.example.com/endpoint" \
  -H "Content-Type: application/json" \
  -d '{"key": "value"}' 2>&1 | tee /tmp/api-debug.log

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 177)May include surrounding context.

bash
# Save full request + response for bug report
curl -v -X POST "https://api.example.com/endpoint" \
  -H "Content-Type: application/json" \
  -d '{"key": "value"}' 2>&1 | tee /tmp/api-debug.log

Static analysis

No suspicious patterns detected.