supermemory

ISSUES FOUND max severity CRITICALSkillSpector Caution 22 registry clawhub.ai last scanned 2026-02-07 1 scan

4 findings (1 critical, 1 high, 2 medium), mainly data exfiltration, command injection, social engineering. Most recent scan 2026-02-07.

This skill presents CRITICAL security risks due to a hardcoded API key exposure and HIGH risks from potential data exfiltration to third-party servers. The hardcoded credential (sm_oiZHA2HcwT4tqSKmA7cCoK_opSRFViNFNxbYqjkjpVNfjSPqQWCNoOBAcxKZkKBfRVVrEQDVxLWHJPvepxqwEPe) must be revoked immediately. The skill encourages storing sensitive data (including API keys) in an external service without adequate warnings about data handling or privacy implications. Additional concerns include command injection risks in bash scripts and misleading documentation that promotes insecure practices. The skill requires immediate remediation before safe use.
Scanner's overall assessment (LLM analyzer)

Findings4

Critical1

High1

Medium2

Low0

Analyzersstatic llm meta

Second scanner: NVIDIA SkillSpector

VerdictCaution 22

Score22 / 100

Issues7

CategoriesData Exfiltration MCP Least Privilege

Version2.12.0 (static)

Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.

What this skill does

Store and retrieve memories using the SuperMemory API. Add content, search memories, and chat with your knowledge base.

Package: 4 files, 6.4 KB
SKILL.md: 255 words, 12+ sections
Code: Shell (3)
Requires: env: SUPERMEMORY_API_KEY

SKILL.md covers: SuperMemory · Setup · Usage · Add a Memory · Add a memory with content · Add a memory with a specific description · Search Memories · Chat with Memories · Implementation · Add Memory · Search Memories · Chat with Memory Base

Findings

CRITICAL Hardcoded API Key Exposed in Documentation Data Exfiltrationllm

The SKILL.md file contains a hardcoded SuperMemory API key (sm_oiZHA2HcwT4tqSKmA7cCoK_opSRFViNFNxbYqjkjpVNfjSPqQWCNoOBAcxKZkKBfRVVrEQDVxLWHJPvepxqwEPe) in the setup instructions. This is a live credential that provides full access to the SuperMemory API account. Anyone with access to this skill can use this key to access, modify, or delete memories, potentially exposing sensitive user data stored in the SuperMemory service.

SKILL.md:9

export SUPERMEMORY_API_KEY="sm_oiZHA2HcwT4tqSKmA7cCoK_opSRFViNFNxbYqjkjpVNfjSPqQWCNoOBAcxKZkKBfRVVrEQDVxLWHJPvepxqwEPe"

Remediation: Immediately revoke the exposed API key through the SuperMemory dashboard. Remove the hardcoded key from documentation. Update instructions to guide users to obtain their own API keys from SuperMemory. Never include live credentials in example code - use placeholder values like 'your-api-key-here' instead.

HIGH Potential Data Exfiltration via Third-Party API Data Exfiltrationllm

The skill sends user content and queries to an external third-party API (api.supermemory.ai) without explicit user consent or data handling disclosure. Users may unknowingly store sensitive information (API keys, passwords, personal data) in their 'memories' which are then transmitted to and stored on external servers. The example usage even suggests storing API credentials: 'Remember that my API key is xyz'. This creates a significant data exposure risk.

scripts/add-memory.sh

Example from SKILL.md: 'Remember that my API key is xyz' → supermemory add 'My API key is xyz'. All scripts make POST requests to https://api.supermemory.ai with user content.

Remediation: Add prominent warnings in SKILL.md about data being sent to external servers. Explicitly warn users NOT to store credentials or sensitive data. Include a privacy/data handling section explaining what data is transmitted and stored externally. Consider implementing local filtering to detect and warn about sensitive data patterns (API keys, passwords) before transmission.

MEDIUM Command Injection Risk in Bash Scripts Command Injectionllm

The bash scripts use user-provided input directly in JSON payloads and shell commands without proper sanitization. While the scripts use double quotes, special characters in user input (like double quotes, backticks, or dollar signs) could break out of the JSON structure or be interpreted by the shell. The CONTENT and QUERY variables are directly interpolated into JSON without escaping.

scripts/add-memory.sh

DATA=$(cat <<EOF
{
    "content": "$CONTENT",
    "customId": "memory_$(date +%s)",
    "containerTags": ["$SANITIZED_DESC"]
}
EOF
)

Remediation: Use proper JSON escaping for user input. Consider using jq or python to construct JSON payloads safely instead of string interpolation. Validate and sanitize input before use. Example: use jq --arg to safely pass variables: jq -n --arg content "$CONTENT" '{content: $content, customId: "memory_\(now)"}'.

MEDIUM Misleading Security Example in Documentation Social Engineeringllm

The documentation provides a dangerous example that encourages users to store API credentials in the memory system: 'Remember that my API key is xyz' → supermemory add 'My API key is xyz' --description 'API credentials'. This is a security anti-pattern that could lead users to store sensitive credentials in an external third-party service, creating credential exposure risks.

SKILL.md

**Store important information:**
- "Remember that my API key is xyz" → `supermemory add "My API key is xyz" --description "API credentials"`

Remediation: Remove examples that suggest storing credentials. Replace with safe examples like storing non-sensitive notes, bookmarks, or learning summaries. Add explicit warnings against storing passwords, API keys, tokens, or other sensitive credentials in the memory system.

What is in the package

File	Kind	Size
`SKILL.md`	doc	2.2 KB
`scripts/add-memory.sh`	script	1.5 KB
`scripts/chat.sh`	script	1.7 KB
`scripts/search.sh`	script	1.0 KB

External hosts referenced

api.supermemory.ai

Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.

Environment variables referenced

API_KEY API_URL SUPERMEMORY_API_KEY MEMORY_ID

Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.

What we checked

Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.

What we did not check

Runtime behaviour. The skill was not executed; findings come from reading its files.
Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
Content the analyzers cannot read: images, binaries, and some non-English text.
Known scanner bypass classes. See the methodology and limitations page.

Sources

Repository: https://clawhub.ai/skills/supermemory
Raw scan result: supermemory-scan.json
Badge for your README: ![skillscan.io](https://skillscan.io/badge/clawhub-ai-supermemory.svg)
Request a re-scan