Back to skill

Security audit

corpus-legal-research

Security checks for vulnerabilities and agentic risk

Overview

This skill is a clearly scoped legal research helper that connects to Corpus for current U.S. law searches, with no bundled executable code or hidden persistence.

Before installing, confirm you are comfortable sending legal research queries to Corpus and, if using the optional Hermes plugin route, prefer a pinned or reviewed plugin revision. Do not use this skill for filings, payments, SSN collection, or legal advice beyond cited research support.

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
README.md:29
Finding

Unpinned Third-Party Plugin Installation Creates a Supply-Chain Risk

Content
View full analysis

Vulnerability Details

File Location: README.md:29-34
Vulnerability Type: Unpinned third-party dependency installation
Risk Level: Medium

Vulnerable Code

markdown
Hermes can also install both skills plus the MCP URL as one Agent Plugins v1
package:

```bash
hermes plugins install teakesdev/corpus-agent-kit/plugins/corpus
hermes plugins enable corpus
text

### Technical Analysis

The documented installation command retrieves a third-party Hermes plugin without specifying an immutable version, release tag, commit hash, cryptographic checksum, or signature. Consequently, the plugin contents installed by this command may change after the reviewed Skill package has been published.

Enabling the plugin immediately after installation increases the risk because any modified code, configuration, or agent instructions supplied by the remote source may become active without an integrity check. This is an insecure dependency practice rather than evidence that the currently referenced plugin is malicious.

### Attack Path

1. An attacker compromises the plugin publisher account, source repository, distribution infrastructure, or another mutable component in the delivery chain.
2. The attacker replaces or modifies the plugin content referenced by `teakesdev/corpus-agent-kit/plugins/corpus`.
3. A user follows the README and executes the unpinned `hermes plugins install` command.
4. Hermes retrieves the attacker-controlled version because the command does not identify an immutable reviewed artifact.
5. The user runs `hermes plugins enable corpus`.
6. Malicious plugin code or instructions execute within the permissions and capabilities available to the Hermes plugin environment.

### Impact Assessment

Successful exploitation could allow attacker-controlled code or agent instructions to operate with the privileges granted to Hermes and the installed plugin. Depending on the host configuration, exposed to
...[truncated 471 chars]
Remediation
View remediation

Remediation Suggestions

  • Pin the plugin installation to an immutable, reviewed release or commit rather than a mutable source reference.
  • Publish the expected plugin version and source revision in the installation documentation.
  • Provide a cryptographic checksum or signed release artifact and require verification before enabling the plugin.
  • Avoid automatically recommending immediate enablement; add an explicit review and integrity-verification step between installation and activation.
  • Document the plugin's required permissions and advise users to grant only the minimum capabilities necessary.
  • Use dependency monitoring and release provenance mechanisms to detect publisher compromise or unexpected artifact changes.
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