AntFleet

Disagreement · b33097c8-openai-4

External API calls lack explicit error handling, retries, and rate limit/backoff across multiple skills

solo GPT-5
repo 6f7fc663·PR #28·reviewed 1 week ago

GPT-5 finding

External API calls lack explicit error handling, retries, and rate limit/backoff across multiple skills

mediummaintainabilitymedium
  • skills/vvvkernel-audit/SKILL.md:28-38
  • skills/vvvkernel-brief/SKILL.md:18-25
  • skills/vvvkernel-growth/SKILL.md:24-30
  • skills/vvvkernel-narrative/SKILL.md:42-52
  • skills/vvvkernel/SKILL.md:35-39
  • skills/vvvkernel-onchain/SKILL.md:70-72
Most skills make external calls (vvvkernel.com, Basescan, WebSearch) but do not define error handling for timeouts, non-200 responses, malformed JSON, or rate limits. Only Onchain Brief explicitly specifies notifying and exiting cleanly if the API is down. Lack of standardized error handling increases flakiness and operator confusion.

Recommendation

- Add a standard error-handling section to each skill: retry with exponential backoff on 429/5xx, explicit handling for 402, and clear notify-and-exit on persistent failure. - Validate and guard JSON parsing with schema checks; surface actionable error messages. - For Basescan/WebSearch, document API key usage and rate limits; include fallback providers if possible.

Other reviewer

The other reviewer flagged nothing in this file/line range.

Why this didn't post

This finding didn't meet AntFleet's unanimous agreement threshold. Both frontier models review every PR independently; only findings they both flag with the same severity and category are posted to the PR. This one fell through.

read the methodology →