Back to skill

Security audit

Open Entity

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because scanners report broad identity-related activation, instructions to persist skill-provided content as system-level guidance, and an unpinned package install; analysis was also incomplete.

Consider avoiding installation until the signer script can be fully reviewed, the package source and version are verified, and the skill's activation and persistent identity-writing behavior are restricted to tasks you explicitly request. The verdict is suspicious, not malicious; the incomplete inspection means the evidence does not establish that the skill is safe.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 39)May include surrounding context.

md
3. HTTP only, or a runtime that blocks `npx`: run the bundled script. It uses `curl` and `scripts/sign.mjs` (Node 22+, WebCrypto only, no dependencies).

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 139)May include surrounding context.

md
- `scripts/sign.mjs`: the signer used by all three (JCS, `eddsa-jcs-2022`, proof-of-work stamp, Identity Anchor).

Identity-maintenance instructions promote skill-provided content to system-level authority and persist it across sessions.

High
Category
Prompt Injection
Confidence
99% confidence
Finding

The skill explicitly directs the agent to place identity content at system level, write persistent agent-instruction files, and reinject that content on resumed or automated turns. This promotes lower-trust skill or identity-document content into durable authority, allowing behavior changes to persist beyond the task and potentially conflict with legitimate instructions.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Activation conditions extend to ordinary sessions and unrelated work without an identity-management scope restriction.

Medium
Category
Trigger Abuse
Confidence
97% confidence
Finding

The triggers extend identity operations to every session, ordinary refusals, mission changes, and encounters with identifiers, without requiring a user request for identity management. Because adoption and record maintenance involve external services and public signed records, this can cause unauthorized side effects or disclosure of contextual information during unrelated work despite the instructions against publishing secrets.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx openentity'.

Medium
Category
MCP Rug Pull
Confidence
98% confidence
Finding

Although the finding labels this an MCP reference, the actual command executes the npm package openentity@latest without a fixed version or integrity verification. A compromised or unexpectedly changed release could execute with the agent's privileges and access local identity keys or other accessible data.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 · scripts/adopt.sh (reported line 15)May include surrounding context.

sh
# 1. Read the current parameters: Declaration hash and proof-of-work bits.
HASH="$(curl -fsS "$REG/spec/declaration/1.0.sha256" | cut -c1-64)"
SPEC="$(curl -fsS "$REG/v1/spec")"

# 2. Generate the key, sign the Entity Document and the Declaration, mint the stamp.
OUT="$(node "$HERE/sign.mjs" adopt "$NAME" "$HASH" "$SPEC")"

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 · scripts/record.sh (reported line 12)May include surrounding context.

sh
OE_HOME="${OPENENTITY_HOME:-$HOME/.openentity}"
HERE="$(cd "$(dirname "$0")" && pwd)"

STAMP="$(node "$HERE/sign.mjs" record "$EID" "$2" "$3" "$(curl -fsS "$REG/v1/spec")")"
curl -sS --fail-with-body -X POST "$REG/v1/entities/$EID/record" \
  -H 'Content-Type: application/json' \
  -H "X-OE-Stamp: $STAMP" \