Install payload
Install payload is mixed and needs spot-checking.
50% (1/2)
Browse the directory
/
Showing 2 resources for "red-teaming"
Saved
/ search · [ ] view density · esc clear input
Active
1 trusted · 1 review in this set — compare to see which signals differ.
Rollout signal scan
1 rollout risk signal in current results
Biggest gaps: metadata review. 0 entries have 2+ required gaps.
watch
Most at-risk entries in this view
Adoption queue
Browse adoption queue · balanced
1/2 visible results are in hold tier and need mitigation before adoption.
ready 0caution 1hold 1
Prompt Injection Defense Guardrails Skill
1 blockers: Metadata review
caution
70/100
Request metadata review from maintainers or internal owners.
skills/prompt-injection-defense-guardrails · trust trusted · confidence 83%
Microsoft PyRIT
2 blockers: Metadata review, Install payload
hold
36/100
Request metadata review from maintainers or internal owners.
Add install/config payload for reproducible team rollout.
Collect package checksum or signed artifact information.
tools/pyrit · trust review · confidence 50%
Decision confidence
Decision confidence scan · balanced
1/2 results are low-confidence and need review before adoption.
high 1medium 0low 1
Prompt Injection Defense Guardrails Skill
Confident candidate for staged adoption.
high
74/100
Missing: Metadata review
skills/prompt-injection-defense-guardrails · trust trusted
Microsoft PyRIT
Hold adoption until Metadata review, Package integrity are resolved.
low
36/100
Missing: Metadata reviewMissing: Package integrityMissing: Install payload
tools/pyrit · trust review
Freshness distribution
50% of this view is aging or stale
Median age 82 days; 1 fresh of 2 scanned. Re-verify the oldest entries.
Oldest entries in this view
Build layered defenses against prompt injection, data exfiltration, and unsafe tool execution in AI agent systems.
Microsoft PyRITby Microsoft · submitted by oktofeesh1
Open-source Python framework from Microsoft for identifying generative AI safety and security risks through automated and human-led red-team assessments.