Back to skill

Security audit

Cerul Openclaw Skill

Security checks for vulnerabilities and agentic risk

Overview

The skill's video-search purpose is clear, but it tells agents to run an unverified remote installer and persist an API key on disk in a way users should review first.

Review this skill before installing. Prefer a pinned, verifiable Cerul CLI installation method, avoid running the curl-to-bash installer automatically, and use a session-scoped or platform-managed CERUL_API_KEY instead of letting the agent print or persist the key in ~/.config/cerul/credentials.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (2)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:26
Finding

Unverified Remote Installer Is Downloaded and Executed by Bash

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:46
Finding

API Credential Is Printed into Agent Output and Mandatorily Persisted in Plaintext

Content
View full analysis
/dev/null ``` ```bash mkdir -p ~/.config/cerul && echo -n "cerul_XXXXX" > ~/.config/cerul/credentials && chmod 600 ~/.config/cerul/credentials ``` ```markdown **Do NOT use `export CERUL_API_KEY=...`** — that only lasts for the current session. Always persist the key to `~/.config/cerul/credentials` so it works across all future sessions and terminal windows. ``` ### Technical Analysis The stated purpose of the first command is to determine whether a credential file exists and is non-empty. Using `cat` exceeds that requirement because it prints the complete secret into tool output. That output may become part of the model context, execution transcript, diagnostic logs, telemetry, or other records accessible beyond the credential file. The Skill also requires the key to be saved across future sessions in a plaintext file and explicitly rejects session-scoped environment use. Persistent authentication can be legitimate for a CLI, and mode `600` limits ordinary access by other local users. However, it does not protect the secret from processes running as the same user, backups, compromised user-level tools, command logging, or disclosure caused by the preceding `cat` command. The metadata identifies `CERUL_API_KEY` as the primary environment credential, making the blanket requirement for plaintext persistence unnecessary for the declared search functionality. The Skill should not choose permanent secret storage on the user's behalf without informed consent. ### Attack Path 1. A Cerul API key already exists in `~/.config/cerul/credentials`, or the user supplies a new key that the agent writes there. 2. During a later invocation, the agent executes `cat ~/.config/cerul/credentials`. 3. The full API key is returned ...[truncated 1199 chars]
Remediation
View remediation
"$HOME/.config/cerul/credentials" ``` 7. Avoid placing the literal secret in an agent-generated shell command because command traces or debugging systems may retain the invocation. 8. Document credential rotation and deletion procedures, and ensure authentication errors never echo the supplied key. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
Findings (9)

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The skill metadata includes a shell install command that pipes a remotely fetched script directly into bash. This executes unverified code from the network at runtime, creating a supply-chain and remote code execution risk if the endpoint, transport, or script content is compromised.

Content

Scanner excerpt · SKILL.md (reported line 26)May include surrounding context.

md
homepage: https://github.com/cerul-ai/cerul
    install:
      - kind: shell
        command: "curl -fsSL https://cli.cerul.ai/install.sh | bash"
        bin: cerul
allowed-tools: Bash(cerul *)
---

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The operational instructions tell the agent to install the CLI by fetching an external script and executing it immediately with bash. In an agent context, this is especially risky because it normalizes autonomous execution of remote code before performing the user's requested task.

Content

Scanner excerpt · SKILL.md (reported line 40)May include surrounding context.

If cerul is not found on PATH, install it first:

bash
curl -fsSL https://cli.cerul.ai/install.sh | bash

Then check if the credentials file exists:

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The use of '| bash' chains network retrieval directly into code execution with no inspection or integrity validation. This pattern is dangerous because it compresses download and execution into one step, making compromise or tampering immediately exploitable.

Content

Scanner excerpt · SKILL.md (reported line 40)May include surrounding context.

If cerul is not found on PATH, install it first:

bash
curl -fsSL https://cli.cerul.ai/install.sh | bash

Then check if the credentials file exists:

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The metadata declares CERUL_API_KEY as the primary environment variable, but the body forbids that safer session-scoped mechanism and mandates writing credentials to disk instead. This inconsistency can bypass platform secret-handling expectations and push the agent toward a less secure storage path.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill explicitly instructs the agent to ask for the user's API key and persist it in a long-lived file for future sessions, which exceeds what is necessary to answer a single video-search request. Persisting secrets on disk increases the chance of unintended reuse, disclosure to other tools/processes, or compromise from later filesystem access.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill tells the agent to save the user's API key in ~/.config/cerul/credentials without any warning about persistence, local exposure, or future-session reuse. Users may provide a secret without understanding it will be stored on disk beyond the immediate task.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

These instructions direct the agent to request the API key from the user, write it to a config file, and keep it available across future sessions. That creates durable secret retention not required by the skill's core video-search function and expands the blast radius if the host is later accessed by other agents, users, or malware.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 52)May include surrounding context.

If the file is empty or missing, ask the user for their API key (get one at https://cerul.ai/dashboard), then save it directly to the config file:

bash
mkdir -p ~/.config/cerul && echo -n "cerul_XXXXX" > ~/.config/cerul/credentials && chmod 600 ~/.config/cerul/credentials

Do NOT use export CERUL_API_KEY=... — that only lasts for the current session. Always persist the key to ~/.config/cerul/credentials so it works across all future sessions and terminal windows.

Session Persistence

Medium
Category
Rogue Agent
Confidence
97% confidence
Finding

The command creates a persistent credentials file and the text explicitly says this is so the key works across future sessions and terminal windows. That is unnecessary retention of a user secret and can enable unauthorized reuse long after the original request is complete.

Content

Scanner excerpt · SKILL.md (reported line 52)May include surrounding context.

If the file is empty or missing, ask the user for their API key (get one at https://cerul.ai/dashboard), then save it directly to the config file:

bash
mkdir -p ~/.config/cerul && echo -n "cerul_XXXXX" > ~/.config/cerul/credentials && chmod 600 ~/.config/cerul/credentials

Do NOT use export CERUL_API_KEY=... — that only lasts for the current session. Always persist the key to ~/.config/cerul/credentials so it works across all future sessions and terminal windows.