Back to skill

Security audit

cross-platform-poster

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly transparent about posting to external services, but it also promotes registering a paid external marketplace service with limited scoping or user-safety guidance.

Review the ClawSwarm marketplace section carefully before installing or using this skill. Only run the posting or registration commands intentionally, use least-privilege and revocable tokens, verify the destination services, and avoid creating a paid public listing unless you understand who operates it, what becomes visible, and how to remove or revoke it.

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
Findings (1)

other

Note
Location
SKILL.md:36
Finding
Unnecessary Third-Party Agent and Paid-Service Registration## Vulnerability Details **File Location**: `SKILL.md`, lines 36–49 **Vulnerability Type**: External Coordination Registration **Risk Level**: Low ### Vulnerable Code ```bash # Join ClawSwarm curl -s -X POST "https://onlyflies.buzz/clawswarm/api/v1/agents/register" \ -H "Content-Type: application/json" \ -d '{"name": "YOUR_NAME", "capabilities": ["social", "posting", "content"]}' # Register cross-posting service (earn HBAR!) curl -s -X POST "https://onlyflies.buzz/clawswarm/api/v1/services/register" \ -H "Authorization: Bearer YOUR_AGENT_ID" \ -H "Content-Type: application/json" \ -d '{"agentId":"YOUR_AGENT_ID","name":"cross_post","description":"Post to Twitter + MoltX + Discord simultaneously","category":"social","pricing":"paid","price_hbar":3}' ``` ### Technical Analysis The skill is presented as a cross-platform content-posting utility, but it also instructs the operator to register an agent and a paid service with the unrelated third-party ClawSwarm marketplace hosted at `onlyflies.buzz`. Executing these commands transmits the selected agent name, advertised capabilities, agent identifier, service description, pricing status, and HBAR price to an external service. The second request uses `YOUR_AGENT_ID` as a bearer credential. Although the instructions do not directly expose a hardcoded secret, operators may provide a real identifier or credential without receiving guidance about its sensitivity, permissions, storage, revocation, or the third party's trust model. This behavior is explicitly documented and requires the operator to execute the commands, so it is not covert code execution. Nevertheless, it expands the skill's scope beyond content formatting and posting and may cause unintended external enrollment or marketplace exposure. ### Attack Path 1. An operator loads the skill expecting cross-platform content-posting functionality. 2. The operator follows the optional “Offer this as a Claw ...[truncated 1261 chars]
Remediation
## Remediation Suggestions 1. Remove the ClawSwarm registration section if external marketplace enrollment is not essential to the skill's core cross-posting purpose. 2. If the integration is retained, place it in a clearly separated, explicitly optional section requiring informed operator consent. 3. Document the third-party operator, privacy policy, transmitted data, retention policy, marketplace visibility, trust assumptions, and account-deletion procedure. 4. Explain whether the returned agent identifier is a credential, what permissions it grants, how it should be stored, and how it can be rotated or revoked. 5. Use a narrowly scoped, revocable authentication token rather than treating a general agent identifier as an unrestricted bearer credential. 6. Warn operators not to insert unrelated social-media API keys, bot tokens, or other platform credentials into these requests. 7. Provide a confirmation or dry-run step that displays all metadata and financial terms before creating the external service. 8. Keep marketplace registration in a separately reviewed integration so use of the core formatting or posting functionality cannot implicitly trigger external enrollment.
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 (3)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill encourages posting content across multiple third-party platforms but does not warn users that both content and platform credentials will be transmitted to external services. This can lead to unintended data disclosure, misuse of privileged tokens, or users authorizing actions they did not fully understand.

External Transmission

Medium
Category
Data Exfiltration
Content
### Post to MoltX
```bash
curl -s -X POST "https://moltx.io/v1/posts" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"content": "Your post here", "tags": ["tag1"]}'
Confidence
91% confidence
Finding
The example performs an authenticated POST to an external service using a bearer API key, which is a genuine external transmission of both user content and sensitive authorization context. In this skill's context, that behavior is expected, but it is still security-relevant because users may expose secrets or publish data externally without sufficient warning or safeguards.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The ClawSwarm instructions cause the agent to register itself and advertise a paid service on an external marketplace, but the skill does not clearly warn the user about this external publication and account-affecting action. That omission increases the risk of unauthorized exposure, reputational impact, and unintended commercial listings tied to the user's identity or agent.

Static analysis

No suspicious patterns detected.