Back to skill

Security audit

AllToken

Security checks for vulnerabilities and agentic risk

Overview

The skill is a mostly coherent AllToken project scaffold, but its HTTP server example can expose chat history and paid API access without authentication or limits.

Install only if you are comfortable sending prompts and media requests to AllToken and using an ALLTOKEN_API_KEY from the environment. Pin dependencies before running the scaffold in a real project. Do not deploy the included HTTP server example unchanged; add authentication, per-user session ownership, rate limits, request validation, retention limits, and avoid returning full conversation history by default.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:86
Finding

Unpinned Third-Party Dependencies Create a Supply-Chain Risk

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:966
Finding

Unauthenticated HTTP Example Exposes Conversation History and Unbounded Session State

Content
View full analysis
(); app.post('/chat', async (req, res) => { const { sessionId, message } = req.body; let agent = sessions.get(sessionId); if (!agent) { agent = createAgent(); sessions.set(sessionId, agent); } res.json({ response: await agent.sendSync(message), history: agent.getMessages() }); }); app.listen(3000); ``` ### Technical Analysis The example exposes a `/chat` endpoint without authentication or authorization. It accepts a caller-controlled `sessionId` and uses that value directly as the key for retrieving an `Agent`. There is no binding between a session and an authenticated principal. The response returns `agent.getMessages()`, disclosing the complete conversation history associated with the supplied identifier. If an attacker learns, predicts, reuses, or otherwise obtains another user's session identifier, the attacker can submit that identifier and retrieve the corresponding conversation history. The `sessions` map also has no capacity limit, expiration, or cleanup mechanism. An attacker can submit many unique session identifiers, causing persistent in-process memory growth. Each accepted message also causes a paid request to the configured remote model API, but the endpoint has no rate limit, request quota, or concurrency control. `express.json()` is enabled without an explicit application-level schema or conservative body limit. Although Express applies a default parser limit, the application does not validate the type or length of `sessionId` and `message`. The example also uses `app.listen(3000)` without explicitly restrictin ...[truncated 2264 chars]
Remediation
View remediation
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 (23)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 747)May include surrounding context.

md
Both Hermes and OpenClaw load skills from `SKILL.md` files and can run TypeScript or Python tools at the agent boundary. There are two integration patterns:

Ssd 3

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The examples explicitly store full user and assistant messages and expose full chat history over an HTTP response without minimization, redaction, retention limits, or access-control discussion. In an agent skill context, conversations often contain secrets, credentials, personal data, or proprietary prompts, so persisting and returning raw transcripts materially increases data leakage risk.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · USAGE.md (reported line 3)May include surrounding context.

md
# Usage — `alltoken` skill

This is the **user-facing manual** for the bootstrap skill at `skills/alltoken/SKILL.md`. If you're an AllToken customer and a teammate just told you to "use the alltoken skill," start here.

## What this skill does

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · USAGE.md (reported line 38)May include surrounding context.

md
# Usage — `alltoken` skill

This is the **user-facing manual** for the bootstrap skill at `skills/alltoken/SKILL.md`. If you're an AllToken customer and a teammate just told you to "use the alltoken skill," start here.

## What this skill does

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · USAGE.md (reported line 114)May include surrounding context.

md
# Usage — `alltoken` skill

This is the **user-facing manual** for the bootstrap skill at `skills/alltoken/SKILL.md`. If you're an AllToken customer and a teammate just told you to "use the alltoken skill," start here.

## What this skill does

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · USAGE.md (reported line 9)May include surrounding context.

md
When loaded by an agent runtime (Hermes, OpenClaw, Claude Code, Codex CLI, OpenCode, Vercel AI SDK chat, …), this skill teaches the host agent how to **bootstrap a new TypeScript or Python project** that talks to AllToken — chat, async image generation, async video generation, model discovery, routing, and cost tracking.

**It is a bootstrap recipe, not a slash-command surface.** Invoking it produces *project files*, not a one-shot API call. If you want direct one-shot commands like `/alltoken-image "a cat"` from the agent prompt, load the sibling skill [`skills/alltoken-call/SKILL.md`](../alltoken-call/SKILL.md) instead.

## Who this is for

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · USAGE.md (reported line 118)May include surrounding context.

md
When loaded by an agent runtime (Hermes, OpenClaw, Claude Code, Codex CLI, OpenCode, Vercel AI SDK chat, …), this skill teaches the host agent how to **bootstrap a new TypeScript or Python project** that talks to AllToken — chat, async image generation, async video generation, model discovery, routing, and cost tracking.

**It is a bootstrap recipe, not a slash-command surface.** Invoking it produces *project files*, not a one-shot API call. If you want direct one-shot commands like `/alltoken-image "a cat"` from the agent prompt, load the sibling skill [`skills/alltoken-call/SKILL.md`](../alltoken-call/SKILL.md) instead.

## Who this is for

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 · SKILL.md (reported line 59)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 144)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 448)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 705)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 725)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 755)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 800)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 805)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 808)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 1049)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 1050)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · SKILL.md (reported line 1051)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · USAGE.md (reported line 53)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · USAGE.md (reported line 68)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

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 · USAGE.md (reported line 81)May include surrounding context.

md
├── package.json
├── tsconfig.json
└── src/
    ├── client.ts        # OpenAI SDK with baseURL=https://api.alltoken.ai/v1
    ├── agent.ts         # Streaming agent core, hooks, tool-loop (Agent class)
    ├── tools.ts         # timeTool, calculatorTool, defaultTools
    ├── media.ts         # generateImage, generateVideo, cancelVideo

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The time tool formats output with toLocaleString('en-US', ...), which forces a specific locale in user-facing output. This is a natural-language/locale policy concern because the skill does not indicate user opt-in or configurability for language/locale.

Content

No source excerpt is available for this finding.