Back to skill

Security audit

LYGO Champion: kairos herald of time

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed persona helper and legacy redirect with narrow local verification helpers, but users should be aware its suggested successor install command is unpinned.

Installers that use `@latest` can change over time, so review the successor skill and consider using a pinned ClawHub version or a sandboxed install if provenance matters. The current artifact itself is advisory and does not show hidden persistence or data access.

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:8
Finding
Unpinned Package Execution and Remote Skill Installation## Vulnerability Details **File Location**: `SKILL.md`, lines 8–9 **Vulnerability Type**: Unpinned third-party package execution and mutable remote Skill installation **Risk Level**: Medium **Complete Code Snippet**: ```markdown > **Consolidated (Δ9 v2):** New installs → `lygo-champion-council`. This slug is legacy retention only. > `npx clawhub@latest install deepseekoracle/lygo-champion-council` ``` ### Technical Analysis The documented command uses `npx` to resolve and execute the latest published version of the `clawhub` npm package. The package version is not pinned, and the remotely installed `deepseekoracle/lygo-champion-council` Skill is not constrained by an immutable version, cryptographic digest, or verified signature. Therefore, the effective code and Skill instructions executed by a user can change after this repository has been audited. Compromise of the npm package, its publisher account, the package registry, or the remote Skill publisher could turn the recommended command into a supply-chain delivery mechanism. The audited repository does not itself contain a malicious payload; the risk arises when a user follows this mutable installation instruction. ### Attack Path 1. An attacker compromises the `clawhub` npm package, its publisher account, or the remote successor Skill. 2. The attacker publishes a malicious release or replaces the remotely resolved Skill content. 3. A user or agent follows the command in `SKILL.md`. 4. `npx` downloads and executes the then-current `clawhub@latest` package. 5. The package installs the remotely resolved `deepseekoracle/lygo-champion-council` Skill. 6. Attacker-controlled package code may execute locally, or attacker-controlled Skill instructions may be loaded in subsequent agent activity. ### Impact Assessment Package code launched through `npx` runs with the privileges of the invoking user. Depending on the malicious upstream payload and loc ...[truncated 411 chars]
Remediation
## Remediation Suggestions 1. Replace `clawhub@latest` with an exact, reviewed package version. 2. Pin the successor Skill to an immutable version or content digest. 3. Publish expected SHA-256 digests or signed provenance and verify them before installation. 4. Use lockfiles and integrity metadata where supported. 5. Prefer downloading the package and Skill into an isolated location for review before executing or activating them. 6. Run installation with least privilege in a sandbox or disposable environment. 7. Document the trusted publisher identity and a procedure for verifying signatures, hashes, and provenance. 8. Periodically review pinned versions before intentionally upgrading them.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding
The description says this skill is only a deprecated slug redirect/legacy identifier for a champion, which implies little or no active behavior. The supplied code instead performs concrete validation logic over several local files: it checks required file existence, parses JSON from references/canon.json, validates the champion field, inspects verifier_usage.md for specific strings, and validates an optional SHA-256 value. That is a materially different primary purpose from a passive deprecated slug. While these are local reads only, the declared permissions list is empty and the behavior is operational self-auditing/verification, not just legacy aliasing. Therefore the description does not accurately represent the code.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The description says this skill is merely a deprecated legacy slug and implies alias/redirect behavior toward another skill, with no substantive function. The supplied code instead performs a concrete action: it accesses a local file, parses JSON, and prints a specific hash value. That is a materially different primary purpose and introduces undeclared file access/output behavior unrelated to a deprecation stub.

Lp3

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding
Without declared permissions the skill's intent is opaque and cannot be validated.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding
The skill instructs users to run 'npx clawhub@latest install ...', which pulls and executes the latest package version at install time. Using a floating latest tag creates a supply-chain risk: a compromised or malicious upstream release could execute unintended code in the user's environment.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding
Suspicious Unicode normalization or mixed-script content

Static analysis

No suspicious patterns detected.