Back to skill

Security audit

retirement-withdrawal-sequencer

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with high confidence; its main caveat is a reported mismatch between advertised Roth conversion modeling and the model’s reported behavior.

The verdict is benign, not risk-free. Before relying on its Roth conversion recommendations, keep in mind that the scanner reports those conversions and their tax and related effects are not included in the year-by-year model; verify projections with a qualified financial or tax professional.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Advertised bracket-fill conversions are recommendations only, not part of the year-by-year model.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The skill advertises modeled bracket-fill conversions, but the simulation never transfers conversion amounts from traditional to Roth accounts or includes their tax and MAGI effects. Nevertheless, it issues actionable conversion recommendations, creating a financial-model integrity risk: users may rely on projections that omit conversion taxes, subsequent RMD changes, and IRMAA consequences. The retirement-planning context makes this consequential; the estimate disclaimer does not resolve the mismatch.

Content

No source excerpt is available for this finding.

Unbounded Output

Medium
Category
Output Handling
Confidence
80% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · scripts/withdrawal_sequencer.py (reported line 184)May include surrounding context.

python
print(f"  Constraints: RMD comes out FIRST (conversion stacks on top); conversions irreversible; "
              f"watch IRMAA tier 1 at {fmt(CONFIG['irmaa'][st][0])}.")
    else:
        print(f"\nRMD YEARS: bracket-fill logic reversed — this year's RMD was {fmt(d_trad)}; "
              f"any conversion stacks on top of it.")
    print(f"\nCash-buffer rule (sequence-of-returns): hold 1-2 yrs of spending "
          f"({fmt(a.spend)}-{fmt(2 * a.spend)}) in cash/short Treasuries so a down market never forces equity sales.")