`ai_safety_disabled`

Ai Safety Disabled — a PullGuard finding type. Findings of this type appear in the PR comment, Step Summary, SARIF (GitHub Security tab / IDE viewers), and the HTML report, each with severity, location, and the remediation guidance below.

How to fix

Re-enable the model safety/content filter; never blanket-disable guardrails low effort

# Before (VULNERABLE): unfiltered output reaches users
safety_settings = {HarmCategory.HARM_CATEGORY_HARASSMENT: HarmBlockThreshold.BLOCK_NONE}

# After (SAFE):
safety_settings = {HarmCategory.HARM_CATEGORY_HARASSMENT: HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE}

// OpenAI: keep moderation on — screen completions before returning them
const flagged = await openai.moderations.create({ input: output });
if (flagged.results[0].flagged) return "I can't help with that.";

Triage

Suppress a confirmed non-issue with a committed .pullguardignore entry (pullguard ignore locally, or comment /pullguard ignore <fingerprint> <reason> on the PR — the fingerprint is printed in the PR comment’s Triage section). Entries support expiresAt for time-boxed snoozes.

Security findings at major or critical severity — and any critical finding — always surface: .pullguardignore cannot hide them. The reviewed paths that keep them visible are acknowledged (reviewed, stays in reports) and, for a confirmed false positive, a reasoned false_positive entry — visible and audited, excluded only from the merge block.