Back to skill

Security audit

Awesome AI Sources

Security checks for vulnerabilities and agentic risk

Overview

This skill is a read-only public RSS feed helper for AI news and related signals, with no hidden credential use or local persistence.

Safe to install for fetching public AI feed summaries. Keep item limits reasonable in detailed mode, especially for the all feed, and only enable recurring delivery if you actually want scheduled fetches.

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
  • 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
Findings (1)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% 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 · SKILL.md (reported line 63)May include surrounding context.

md
- `headlines` — title only. Compact list, just "what happened."
   - `summary` — title + the AI-generated summary / overview / engagement stats (whichever the feed provides) + the source link. The default Agentic Brew item shape.
   - `detailed` — title + full description (no truncation) + source link + any `content:encoded` inner content (e.g., tweet list for twitter, overview bullets for news) + the `<category>` tags.

   To apply the choice: fetch with `--json` internally, then format the items per the chosen detail level. Do NOT pass `--limit` so low that you lose information the user asked for — only `--limit` controls *how many* items, not how *deep* each one goes.