Back to skill

Security audit

OpenClaw Verified Upgrade

Security checks for vulnerabilities and agentic risk

Overview

This skill provides a controlled OpenClaw upgrade workflow with verified backups and clear limits on sensitive changes.

Use this skill when you intend to authorize software upgrades, sensitive local backups and service interruptions. Specify the target version, recovery limits and backup retention, and protect backups containing credentials. Follow the current workflow rather than the legacy script. Manual review addressed the scanner's material coverage gap; compatibility and actual upgrade success still depend on the installed host and selected release.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Memory Manipulation

High
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.

Content

Scanner excerpt · SKILL.md (reported line 100)May include surrounding context.

md
Target installed plus later failure means post-update recovery, not another core install. Inspect plugin compatibility, convergence, integrity and capability consent before blaming Node/cache/OAuth. update repair can mutate Doctor/config/plugin state; it is not a status command.

Never automatically restore databases/config, force-install services, reset credentials, disable plugins or purge state as a recovery shortcut. Review official compatibility-checked package recovery against the user's policy before admission and report its actual outcome.

Report success / already-current / installed-but-unverified / rolled-back / blocked / failed, from -> target -> actual runtime, native or external case/run IDs, backups, semantic differences, critical checks, observation evidence, cleanup and remaining action. Terminal updater success alone is insufficient.

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

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.

Session Persistence

Medium
Category
Rogue Agent
Confidence
65% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · examples/test-prompts.json (reported line 42)May include surrounding context.

json
"id": "no-surviving-transport",
    "prompt": "Run the update in the background and carry on after restart.",
    "context": "The shell is a Gateway descendant; no native handoff/coordinator or genuine external runner exists.",
    "expected": "Block before core mutation, identify missing external execution; do not claim nohup or an exec ID proves survival."
  },
  {
    "id": "handoff-accepted",

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · examples/test-prompts.json (reported line 101)May include surrounding context.

json
{
    "id": "private-runtime-phases",
    "prompt": "Preserve system Node and finish the approved runtime/core phases.",
    "context": "The new private runtime is verified. The plist points to it, but the live old-host process still uses old Node after a kickstart. Controlled reload in this phase is approved.",
    "expected": "Verify actual process and use the supported approved reload, then validate old host under new Node and refresh backup before core; do not rely on plist/CLI or impose an unrelated one-action counter."
  },
  {

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · examples/test-prompts.json (reported line 102)May include surrounding context.

json
"id": "private-runtime-phases",
    "prompt": "Preserve system Node and finish the approved runtime/core phases.",
    "context": "The new private runtime is verified. The plist points to it, but the live old-host process still uses old Node after a kickstart. Controlled reload in this phase is approved.",
    "expected": "Verify actual process and use the supported approved reload, then validate old host under new Node and refresh backup before core; do not rely on plist/CLI or impose an unrelated one-action counter."
  },
  {
    "id": "slow-start",

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · references/migration-and-runtime.md (reported line 31)May include surrounding context.

md
Discover target runtime requirements and the installation's supported private-runtime discovery locations. Verify official source/checksum and architecture. Do not hardcode one historical Node version or change system Node/shell profiles when a private runtime was selected.

Record both shell CLI and actual service/Gateway executable. A launcher can retry into a private Node while launchd still holds cached old arguments. Reading a changed plist or running CLI --version alone does not prove the service switched.

For an authorized multi-stage plan:
1. Install the verified private runtime and change only reviewed runtime/service fields.

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE.md (reported line 7)May include surrounding context.

md
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.