ClawHub
Back to skill

Security audit

Spaced Repetition

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only learning guide for spaced repetition and does not contain code, hidden actions, credential handling, or persistence.

This skill appears safe to install as an educational reference. It may recommend workflow changes such as deleting or suspending bad flashcards and installing related skills only after confirmation, so users should treat those as learning-advice decisions rather than automatic actions.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
85% confidence
Finding

Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Content

Scanner excerpt · SKILL.md (reported line 37)May include surrounding context.

md
- SM-2 (Anki's legacy engine, from SuperMemo, 1987): each grade multiplies the interval by an ease factor; ease drifts up on "easy", down on "hard". No target retention; the schedule is a heuristic.
- SM-2 weakness: ease inflates on early easy cards then never recovers, producing absurd intervals (a card "known" after three easy taps lands at a 6-month interval you will actually fail). Interval explosion is the tell.
- FSRS (Free Spaced Repetition Scheduler): optimizes to a desired retention (default 0.9) via 19 parameters fit to your personal review history. Replaces SM-2 in Anki as of ~2024; strictly better once you have ~1k reviews logged to fit on.
- FSRS needs history: on a fresh profile with <1000 reviews it falls back to defaults; the fit improves as data accumulates. Do not judge FSRS on the first week.
- Set desired retention to 0.9 unless you have a reason. 0.85 = fewer reviews, more forgetting; 0.95+ = review load balloons. 0.97 costs ~2x the reviews of 0.90 (FSRS).
- Retention target is the lever, not the interval: decide how much you can afford to forget, set the target, let the algorithm produce the intervals. Hand-tuning intervals per card is the #1000's rabbit hole.
- Re-grade on actual recall, not difficulty feeling: marking "hard" because you disliked the question inflates load. Mark hard only when recall was slow or partial, not when the card annoys you.