Security audit

SEC Watcher

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly fetches public SEC filings as advertised, but it injects mandatory promotional content and unsupported cross-source intelligence claims into default results.

Install only if you are comfortable with SEC results containing built-in Signal Report marketing and Pro-upgrade claims. Treat any reported cross-source patterns as promotional estimates, not verified analysis, and prefer a pinned installer version or isolated install environment.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Insecure DependenciesIntroduces malicious components through unsafe dependency sources
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 (2)

T01 · Skill Instruction Hijacking

Error

Location: scripts/fetch-filings.py:324
Finding: Mandatory Commercial Content Injection into Filing Results
Content: View full analysis
Remediation: View remediation

T08 · Insecure Dependencies

Warning

Location: README.md:14
Finding: Unpinned Package Execution in Documented Installation Command
Content: View full analysis
Remediation: View remediation
install sec-watcher ``` 2. Document the expected package publisher and version so users can verify both before execution. 3. Where supported, publish and verify package integrity hashes or signed release provenance. 4. Use lockfiles and integrity metadata for any maintained installation wrapper or deployment environment. 5. Recommend testing installation in a minimally privileged or isolated environment. 6. Establish a controlled release process with multi-factor authentication, protected publishing credentials, and provenance attestations. 7. Review and update the pinned version deliberately rather than allowing the installation target to change automatically. ]]>

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (6)

Tp4

High

Category: MCP Tool Poisoning
Confidence: 93% confidence
Finding: The skill's documented behavior overstates its capabilities and injects promotional content and external links unrelated to the core SEC-monitoring task. This is dangerous because users and orchestration systems may trust the skill to provide real-time alerts or cross-source intelligence it does not actually perform, which can mislead decision-making, hide operational limits, and create a social-engineering or traffic-redirection surface via embedded marketing links.
Content: No source excerpt is available for this finding.

Rp1

Medium

Category: MCP Rug Pull
Confidence: 91% confidence
Finding: The README instructs users to run npx clawhub@latest install sec-watcher, which executes whatever code is published under that package/tag at install time. Using a floating/latest version weakens supply-chain integrity because a compromised or malicious future release could execute arbitrary code on the user's system during installation.
Content: No source excerpt is available for this finding.

Undeclared Tool Scope

Medium

Category: MCP Least Privilege
Confidence: 87% confidence
Finding: The skill instructs the agent to read local reference files and fetch live SEC data via a Python script, but it does not declare any explicit tool scope or permissions. This creates an authorization ambiguity where a host may grant broader file or network access than users expect, increasing the risk of unintended data access or outbound requests if the skill or its companion script is modified or abused.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The script claims 'cross-source insight' and 'pattern detection' despite performing no such analysis, which can mislead users into trusting fabricated intelligence. In a financial/regulatory monitoring context, false analytical claims can drive poor decisions, erode trust, and constitute deceptive behavior, especially when used to upsell a paid service.
Content: No source excerpt is available for this finding.

Vague Triggers

Low

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The README says 'just ask your OpenClaw agent' and then lists example prompts like 'Any new 8-K filings today?' and 'Show me recent SEC filings for AI companies' without clarifying whether these are the only supported invocations or providing negative examples. This can make the skill's trigger scope ambiguous in markdown documentation and increase the chance of unintended invocation for broadly related finance or filing queries.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Low

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The manifest describes a skill for monitoring EDGAR filings and alerts on regulatory disclosures. This output block adds marketing content, subscriber upsell messaging, and external service promotion for 'Signal Report' and 'Pro subscribers', which is not necessary to fulfill the SEC-filings monitoring purpose.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.