Skip to main content

Browse the directory

Showing 4 resources for "red-teaming"
Saved
Active

1 trusted · 3 review in this set — compare to see which signals differ.

Trust snapshot

4 results in this view

Claimed
0%(0/4)

3 trust signals differ in this sample: Package trust, Source provenance, Submitter

Signals differ on Package trust, Source provenance, Submitter — add entries to compare before you install.

Rollout signal scan

3 rollout risk signals in current results

Biggest gaps: metadata review, package integrity. 2 entries have 2+ required gaps.

4 scanned

Install payload

Install payload is sparse; verify before rollout decisions.

risk

25% (1/4)

Adoption queue

Browse adoption queue · balanced

3/4 visible results are in hold tier and need mitigation before adoption.

ready 0caution 1hold 3
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%

Promptfoo

3 blockers: Metadata review, Safety notes

hold

22/100

Request metadata review from maintainers or internal owners.

Capture safety notes with misuse/guardrail guidance.

Add install/config payload for reproducible team rollout.

tools/promptfoo · trust review · confidence 33%

Garak

3 blockers: Metadata review, Safety notes

hold

10/100

Request metadata review from maintainers or internal owners.

Capture safety notes with misuse/guardrail guidance.

Add install/config payload for reproducible team rollout.

tools/garak · trust review · confidence 17%

Decision confidence

Decision confidence scan · balanced

3/4 results are low-confidence and need review before adoption.

high 1medium 0low 3

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

Promptfoo

Hold adoption until Metadata review, Safety notes are resolved.

low

22/100

Missing: Metadata reviewMissing: Safety notesMissing: Package integrity

tools/promptfoo · trust review

Garak

Hold adoption until Metadata review, Safety notes are resolved.

low

10/100

Missing: Metadata reviewMissing: Safety notesMissing: Privacy notes

tools/garak · trust review

Freshness distribution

75% of this view is aging or stale

Median age 92 days; 1 fresh of 4 scanned. Re-verify the oldest entries.

median 92d

Aging

91–180 days

75%

3 entries

Stale

> 180 days

0%

0 entries

Theme distribution

Results center on security

75% of this view shares the top theme. Leading themes: security, red-teaming, ai-red-teaming.

Focused

9 distinct themes across 4 scanned

Promptfoo logo

Open-source prompt testing and red-teaming framework for LLM outputs, regressions, evaluations, and security checks.

Safety · Privacy ✓

Build layered defenses against prompt injection, data exfiltration, and unsafe tool execution in AI agent systems.

Level:advancedType:generalVerified:draft
Safety ✓ Privacy ✓

Open-source LLM vulnerability scanner for probing model behavior, prompt attack surfaces, and safety failures.

Safety · Privacy ·
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.