Skill Trust Decision

polymarket-pro

The skill documents legitimate Polymarket trading functionality but includes a dangerous curl|bash installation method that could be exploited if the remote repository is compromised, and inadequately warns about private key handling risks.

Install decision first Source: ClawHub Scanned: Jun 24, 2026
Files 3
Artifacts 3
Violations 0
Findings 4
Most direct threat evidence
01
User follows documented curl|sh installation from GitHub raw content supply_chain · SKILL.md
02
If repo is compromised or MITM occurs, attacker injects malicious code into install.sh supply_chain · https://raw.githubusercontent.com/Polymarket/polymarket-cli/main/install.sh
03
Malicious code executes with user's privileges - could install backdoors, steal keys, exfiltrate data Impact · User's shell

Why this conclusion was reached

2/4 dimensions flagged
Pass
Declared vs actual capability

Declared resources and inferred behavior are broadly aligned.

Block
Hidden execution and egress

1 high-risk artifacts or egress signals were extracted.

Block
Attack chain and severe findings

The report includes 3 attack-chain steps and 1 severe findings.

Review
Dependencies and supply chain hygiene

1 dependency or supply-chain issues need attention.

Attack Chain

01
User follows documented curl|sh installation from GitHub raw content

supply_chain · SKILL.md:35

02
If repo is compromised or MITM occurs, attacker injects malicious code into install.sh

supply_chain · https://raw.githubusercontent.com/Polymarket/polymarket-cli/main/install.sh:1

03
Malicious code executes with user's privileges - could install backdoors, steal keys, exfiltrate data

Impact · User's shell

What drove the risk score up

Dangerous curl|bash installation +20

Line 35 documents curl -sSL https://raw.githubusercontent.com/Polymarket/polymarket-cli/main/install.sh | sh - a classic supply chain attack vector if repo is compromised

Unprotected private key handling +12

Documents CLI flag --private-key which exposes keys in process list and shell history

No allowed-tools declaration +5

SKILL.md lacks allowed-tools specification in frontmatter

Remote script execution documented +5

Shell script installation path is documented without version pinning or checksum verification

Most important evidence

High Supply Chain

Dangerous curl|bash installation documented

SKILL.md line 35 documents a curl|bash installation pattern that downloads and executes remote script without verification. If the GitHub repository is compromised or a MITM attack occurs, arbitrary code could execute on the user's system.

SKILL.md:35
Remove this installation method. Use only Homebrew (which has code signing) or direct binary download with SHA256 checksum verification. Document the specific version to install.
Medium Credential Theft

Insecure private key handling documented

The skill documents passing private keys via CLI flag (--private-key 0xabc...) which exposes keys in process list (ps aux), shell history, and log files. This is a high risk for credential theft if the system is compromised.

SKILL.md:41
Strongly recommend against CLI flag approach. Emphasize environment variable or config file methods with restricted permissions (chmod 600). Suggest users consider hardware wallet integration if available.
Medium Doc Mismatch

allowed-tools not declared in SKILL.md

The skill frontmatter lacks allowed-tools specification, making it unclear what tools the AI agent is permitted to use when executing this skill.

SKILL.md:1
Add allowed-tools specification: Read (for config files), Bash (for polymarket CLI), potentially Write (for config creation). Document these in the skill frontmatter.
Low Supply Chain

Installation dependencies not version-pinned

The skill requires 'polymarket CLI (brew install or shell script)' and 'jq' without specifying versions. Homebrew formulas can change over time.

SKILL.md:10
Specify minimum versions or commit SHAs for reproducibility. Document known-working version combinations.

Declared capability vs actual capability

Shell Pass
Declared WRITE
Inferred WRITE
SKILL.md documents polymarket CLI execution extensively
Network Pass
Declared READ
Inferred READ
CLOB API calls, market data retrieval documented
Filesystem Pass
Declared NONE
Inferred WRITE
Config file writes (~/.config/polymarket/config.json) are necessary for wallet storage
Environment Pass
Declared NONE
Inferred READ
POLYMARKET_PRIVATE_KEY access is documented and necessary for trading

Suspicious artifacts and egress

Critical Dangerous Command
curl -sSL https://raw.githubusercontent.com/Polymarket/polymarket-cli/main/install.sh | sh

SKILL.md:35

Medium External URL
https://clawhub.ai/user/liwagu

skill-card.md:7

Medium External URL
https://clawhub.ai/liwagu/polymarket-pro

skill-card.md:28

Dependencies and supply chain

PackageVersionSourceKnown vulnNotes
polymarket-cli unspecified brew or GitHub No Version not pinned; curl|sh installation is a supply chain risk
jq unspecified brew/apt No Standard JSON processing tool, commonly pre-installed

File composition

3 files · 542 lines
Markdown 2 files · 537 linesJSON 1 files · 5 lines
Files of concern · 2
SKILL.md Markdown · 494 lines
Dangerous curl|bash installation documented · Insecure private key handling documented · allowed-tools not declared in SKILL.md · Installation dependencies not version-pinned · curl -sSL https://raw.githubusercontent.com/Polymarket/polymarket-cli/main/install.sh | sh
skill-card.md Markdown · 43 lines
https://clawhub.ai/user/liwagu · https://clawhub.ai/liwagu/polymarket-pro
Other files · _meta.json

Security positives

skill-card.md includes thoughtful 'Known Risks and Mitigations' section acknowledging the curl|bash danger
The core functionality appears legitimate - Polymarket trading via official CLI
No evidence of obfuscation, base64 execution, or hidden functionality
Documents both JSON and table output formats for transparency
No credential exfiltration or suspicious network calls beyond expected API usage
Dependencies (jq, polymarket CLI) are standard, well-known tools