Back to skill

Security audit

Remix V2 Error Boundaries Review

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Remix code-review skill, but it requires following an unbundled external verification protocol that is outside the reviewed package.

Review or vendor the external `review-verification-protocol` dependency before relying on this skill. The Remix guidance itself is scoped and non-executable, but the skill's final reporting process depends on instructions outside the reviewed artifact.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:115
Finding

Mandatory Unpinned External Skill Instruction Dependency

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 115-116
Vulnerability Type: Untrusted external instruction dependency
Risk Level: Medium

Vulnerable Code

markdown
4. **Protocol** — **Pass:** You completed the Pre-Report Verification
   Checklist in [review-verification-protocol](../../../beagle-core/skills/review-verification-protocol/SKILL.md)

Technical Analysis

The Skill makes completion of its reporting process conditional on following instructions from ../../../beagle-core/skills/review-verification-protocol/SKILL.md, which is outside the audited project package. The referenced content is not bundled with this Skill, pinned to an immutable version, or protected by an integrity check.

Consequently, the effective instructions executed by an agent cannot be established by reviewing this package alone. A party able to modify or replace the sibling Skill could introduce instructions that change audit criteria, suppress findings, manipulate output, or direct additional tool use. The audit did not inspect that external file, so this finding does not assert that its current contents are malicious; the vulnerability is the mandatory trust placed in mutable, unverified instructions.

Attack Path

  1. An attacker gains write access to the sibling beagle-core/skills/review-verification-protocol/SKILL.md file or controls how that dependency is provisioned.
  2. The attacker inserts instructions that manipulate the agent’s review process, such as suppressing selected findings or requesting unrelated tool operations.
  3. An agent loads the audited Skill and reaches the mandatory protocol hard gate.
  4. To satisfy that gate, the agent reads and follows the externally controlled checklist.
  5. The injected instructions influence the audit session or its final report.

Impact Assessment

Successful exploitation could compromise the integrity and completeness of audits performed with this Skil ...[truncated 398 chars]

Remediation
View remediation

Remediation Suggestions

  1. Vendor the required verification checklist into this project so the complete effective instruction set is reviewable as one package.
  2. If an external dependency is unavoidable, pin it to an immutable, reviewed version and verify its cryptographic digest before use.
  3. Remove the mandatory hard gate on mutable content outside the project boundary.
  4. Explicitly state that referenced material is untrusted and cannot override system or user instructions, safety controls, approved task scope, or reporting integrity.
  5. Include the external dependency in release-time security reviews and fail closed when its identity or integrity cannot be verified.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 41)May include surrounding context.

md
ow 4xx/5xx, swallowing `error.data`, throwing strings, missing `handleError` | [references/throw-response.md](references/throw-response.md) |

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 139)May include surrounding context.

md
ow 4xx/5xx, swallowing `error.data`, throwing strings, missing `handleError` | [references/throw-response.md](references/throw-response.md) |

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 116)May include surrounding context.

md
finding is about v2 correctness.

4. **Protocol** — **Pass:** You completed the Pre-Report Verification
   Checklist in [review-verification-protocol](../../../beagle-core/skills/review-verification-protocol/SKILL.md)
   for this review.

## Review Questions

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The canonical example specifies <html lang="en">, which enforces English as the document language in the sample implementation. The file does not indicate that this is optional, user-selectable, or justified as a region-specific constraint, so it can conflict with organizational language/locale choice policies.

Content

No source excerpt is available for this finding.