ClawHub
Back to skill

Security audit

keyword-scout

Security checks for vulnerabilities and agentic risk

Overview

This SEO keyword skill is mostly purpose-aligned, but it asks the agent to run an unbundled local validator from a dynamically found skills root.

Install only if you trust the surrounding Groundcrew skills installation and can verify the validator path, ownership, and integrity. Expect the skill to use user-provided exports or configured SEO providers, and review any paid API batch before approval.

A.I.G

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:7
Finding

Execution of an Unverified External Local Validator

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, line 7
Vulnerability Type: T09: Insecure Skill Coding Practices
Risk Level: Medium

Vulnerable code snippet:

text
Locate the active skills root and run `<skills-root>/.groundcrew/groundcrew-doctor.py --evidence <record.json>` before using them in a conclusion.

Technical Analysis

The skill requires the agent to locate an “active skills root” dynamically and execute groundcrew-doctor.py from that location. The referenced program is not included in the audited project, and the instruction does not define:

  • A canonical, trusted installation path
  • Expected file ownership or permissions
  • A cryptographic hash or signature
  • A trusted package or release source
  • Rejection of user-writable or otherwise untrusted roots

Consequently, the security of the execution depends on an external filesystem location that was not part of this audit. If root discovery selects an attacker-controlled or writable directory, a malicious file placed at .groundcrew/groundcrew-doctor.py could be executed as though it were the legitimate validator.

This is an insecure external-tool invocation rather than confirmed embedded malicious code: the audited project contains no executable scripts or malicious payload.

Attack Path

  1. An attacker gains write access to a directory that can be selected as the active skills root.
  2. The attacker creates or replaces .groundcrew/groundcrew-doctor.py with a malicious Python program.
  3. The skill processes an evidence record and directs the agent to run the validator.
  4. The agent resolves the ambiguous skills root to the attacker-controlled directory.
  5. The malicious program executes with the privileges and environment available to the agent process.
  6. The program can access resources available under those privileges, subject to operating-system and sandbox restrictions.

Impact Assessment

Successful exploitation could provide arbitrary local code execu ...[truncated 485 chars]

Remediation
View remediation

Remediation Suggestions

  1. Bundle the validator in the reviewed skill package or install it through a fixed, trusted distribution mechanism.
  2. Resolve the validator through a canonical path rather than dynamically searching for an active skills root.
  3. Verify the validator’s cryptographic hash or signature against a pinned trusted value before every execution.
  4. Check that the validator and all parent directories have trusted ownership and are not writable by untrusted users.
  5. Refuse execution when the validator is absent, duplicated across candidate roots, symlinked outside the trusted installation, or fails integrity checks.
  6. Invoke the interpreter and arguments directly without shell interpolation.
  7. Validate and canonicalize the evidence-record path, and pass it as a discrete argument.
  8. Run the validator with minimal filesystem, environment, credential, and network access.
  9. Document the trusted source, supported version, expected permissions, and fail-closed behavior.

SkillSpector

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 (1)

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 · references/connectors.md (reported line 48)May include surrounding context.

md
- **Tier 1.** Google Suggest for expansion (unofficial endpoint; treat as hints, `confidence: low`). A SERP the user pastes or fetches is an observation for that query, locale, device, and date — record all four.
- **DataForSEO.** Paid, per-request. Read `dataforseo.md`; show the cost preflight and get approval for each batch.
- **Ahrefs MCP** (`https://api.ahrefs.com/mcp/mcp`, API key on a plan that includes MCP) and **Semrush MCP** (`https://mcp.semrush.com/v1/mcp`, OAuth or API key). Use only when the user already pays for the tool and has configured the MCP in their client. Label rows `source: "ahrefs"` / `"semrush"`, `confidence` per the vendor's own accuracy notes; these are third-party estimates, not the user's own data.
- **TrustGrowth.** `keywords` (typed opportunities with volume, difficulty, intent, current position, scores) for the verified site. SERP snapshots per keyword are scheduled (TEC-7796).

### `~~link database` — backlinks and authority

Static analysis

No suspicious patterns detected.