Back to skill

Security audit

auth-md-signup

Security checks for vulnerabilities and agentic risk

Overview

The skill’s signup flow is coherent and consent-based, but its optional persistent token store can save active bearer tokens and email addresses without mandatory local file protections.

Review this before installing if you plan to persist tokens. Leaving AUTH_MD_TOKEN_STORE unset keeps tokens in session memory only; if you enable it, use a private file or secret store with strict permissions and revoke tokens if that file may have been exposed.

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:222
Finding

Persistent Bearer Token Store Lacks Mandatory File Access Controls

Content
View full analysis
", "email": "", "registered_at": "" }` If `AUTH_MD_TOKEN_STORE` env var is set, append to that JSON file. Otherwise, hold in session memory only. ``` ### Technical Analysis The Skill instructs the Agent to persist an active bearer credential and the user's email address in a JSON file selected through `AUTH_MD_TOKEN_STORE`. It does not require the file to be created with restrictive permissions, verify that the destination is owned by the current user, reject symbolic links, or use an atomic write procedure. Because the environment variable may specify an arbitrary path, the destination could be an existing file with permissive permissions or a symbolic link controlled by another local process or user. A bearer token can generally be used by anyone who obtains it, without additional proof of identity. Although the token is service-scoped, plaintext disclosure would permit access within all scopes granted to that token. The network transmission of the email, OTP, claim token, and bearer token is otherwise consistent with the declared signup and revocation workflow. The Skill also requires explicit user consent and endpoint-domain validation. The confirmed issue is limited to insecure local persistence rather than unauthorized network exfiltration. ### Attack Path 1. An attacker or compromised local process creates a destination file readable by another account, or creates a symbolic link to an attacker-accessible destination. 2. `AUTH_MD_TOKEN_STORE` is configured to reference that destination, or an existing token-store file has overly permissive permissions. 3. The user completes the auth.md registration flow. 4. The Agent writes the active service t ...[truncated 1055 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · README.md (reported line 22)May include surrounding context.

md
---
name: auth-md-signup
description: Complete an auth.md USER-CLAIMED signup (OTP flow) against any service that publishes the auth.md protocol. Consent-preserving — the human confirms the OTP; the agent never auto-confirms. Reference implementation targets NoForm (noform.dev).
version: 1.0.0
metadata:
  openclaw:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

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

md
---
name: auth-md-signup
description: Complete an auth.md USER-CLAIMED signup (OTP flow) against any service that publishes the auth.md protocol. Consent-preserving — the human confirms the OTP; the agent never auto-confirms. Reference implementation targets NoForm (noform.dev).
version: 1.0.0
metadata:
  openclaw:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 29)May include surrounding context.

md
---
name: auth-md-signup
description: Complete an auth.md USER-CLAIMED signup (OTP flow) against any service that publishes the auth.md protocol. Consent-preserving — the human confirms the OTP; the agent never auto-confirms. Reference implementation targets NoForm (noform.dev).
version: 1.0.0
metadata:
  openclaw:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 230)May include surrounding context.

md
---
name: auth-md-signup
description: Complete an auth.md USER-CLAIMED signup (OTP flow) against any service that publishes the auth.md protocol. Consent-preserving — the human confirms the OTP; the agent never auto-confirms. Reference implementation targets NoForm (noform.dev).
version: 1.0.0
metadata:
  openclaw:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 309)May include surrounding context.

md
---
name: auth-md-signup
description: Complete an auth.md USER-CLAIMED signup (OTP flow) against any service that publishes the auth.md protocol. Consent-preserving — the human confirms the OTP; the agent never auto-confirms. Reference implementation targets NoForm (noform.dev).
version: 1.0.0
metadata:
  openclaw:

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 150)May include surrounding context.

Step 6 — Anonymous registration

bash
curl -sS -X POST "<identity_endpoint>" \
  -H "Content-Type: application/json" \
  -d '{"type": "anonymous"}'

Static analysis

No suspicious patterns detected.