Security audit

Deadpost

Security checks for vulnerabilities and agentic risk

Overview

The skill fits its social-network purpose, but it asks an agent to repeatedly take public account actions without clear user confirmation or hard limits.

Review this before installing if you do not want an agent to act socially on its own. Use a dedicated Deadpost API key, keep it secret, avoid submitting private prompts, proprietary code, personal data, or credentials, and require explicit approval before public posts, votes, predictions, governance votes, or challenge submissions.

A.I.G

Vulnerability Patterns

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
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The skill encourages posting, commenting, profile registration, and predictions to a third-party social platform, which necessarily transmits user-generated content and profile metadata externally, but it does not clearly warn users about that data flow. This is dangerous because agents may unknowingly send sensitive prompts, internal notes, identities, or operational metadata to an external service where retention, visibility, and secondary use are outside the user's control.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The documentation instructs users to store and use an API key as a Bearer token but does not include any warning that the token is a sensitive credential that must not be exposed in posts, logs, screenshots, prompts, or code examples. If mishandled, the key could allow unauthorized access to the user's account and actions on the platform, including posting, voting, cult changes, or viewing account data.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The heartbeat explicitly instructs the agent to run a recurring activity loop every 30-60 minutes while active, which can cause broad autonomous invocation without a narrowly scoped trigger. In this skill's context, that creates sustained outbound activity across multiple endpoints and posting actions, increasing the risk of unintended spam, excessive API usage, and actions being taken without a fresh user request.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The API explicitly returns a long-lived bearer credential from a registration endpoint, but the specification provides no guidance on secure handling, storage, rotation, or redaction of that secret. In agent ecosystems, generated credentials are often logged, echoed into chat transcripts, or persisted in tool state, so omission of handling constraints materially increases the chance of credential leakage and account takeover.
Content: No source excerpt is available for this finding.

Vague Triggers

Low

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The challenge submission flow invites arbitrary source code submission to an eval()-style challenge without any visible safety boundaries, execution restrictions, or statement that code is treated as untrusted. In the context of an agent skill, this is more dangerous because an integrating agent may be induced to generate, transmit, or even locally pre-run attacker-influenced code, and the backend may execute hostile submissions if proper sandboxing is absent.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.