Returns the built-in policy names and their high-level behavior. Use these
names directly in scanners and orchestration helpers, for example
scan_prompt("text", policy = "comprehensive").
Value
A data frame with policy names, descriptions, rule counts,
thresholds, rate-guard availability, and a selected column when requested.
Examples
available_policies()
#> name
#> 1 enterprise_default
#> 2 baseline
#> 3 pharma_gxp
#> 4 finance_strict
#> 5 education_safe
#> 6 open_research
#> 7 comprehensive
#> 8 custom
#> description
#> 1 Production baseline for prompt injection, NLP intent, PII/PHI, secrets, system-prompt extraction, and agency language.
#> 2 Backward-compatible alias for enterprise_default.
#> 3 Clinical and regulated-workflow controls on top of enterprise_default.
#> 4 Financial-services controls with account, advice, trading, and rate-guard checks.
#> 5 Education controls for minor PII and academic-integrity bypasses.
#> 6 Lighter research profile focused on injection and secrets with higher thresholds.
#> 7 Maximum built-in coverage across enterprise, clinical, finance, education, code, and rate limits. Uses moderate thresholds (redact_at = 0.4, block_at = 0.7). For pharma-tier strictness, supply overrides = list(thresholds = list(redact_at = 0.3, block_at = 0.6)) explicitly.
#> 8 Empty policy for fully custom rule sets.
#> rules redact_at block_at rate_guard
#> 1 14 0.4 0.75 FALSE
#> 2 14 0.4 0.75 FALSE
#> 3 18 0.3 0.60 FALSE
#> 4 17 0.4 0.75 TRUE
#> 5 16 0.4 0.75 FALSE
#> 6 8 0.8 0.95 FALSE
#> 7 23 0.4 0.70 TRUE
#> 8 0 0.4 0.75 FALSE
available_policies("comprehensive")
#> name
#> 1 enterprise_default
#> 2 baseline
#> 3 pharma_gxp
#> 4 finance_strict
#> 5 education_safe
#> 6 open_research
#> 7 comprehensive
#> 8 custom
#> description
#> 1 Production baseline for prompt injection, NLP intent, PII/PHI, secrets, system-prompt extraction, and agency language.
#> 2 Backward-compatible alias for enterprise_default.
#> 3 Clinical and regulated-workflow controls on top of enterprise_default.
#> 4 Financial-services controls with account, advice, trading, and rate-guard checks.
#> 5 Education controls for minor PII and academic-integrity bypasses.
#> 6 Lighter research profile focused on injection and secrets with higher thresholds.
#> 7 Maximum built-in coverage across enterprise, clinical, finance, education, code, and rate limits. Uses moderate thresholds (redact_at = 0.4, block_at = 0.7). For pharma-tier strictness, supply overrides = list(thresholds = list(redact_at = 0.3, block_at = 0.6)) explicitly.
#> 8 Empty policy for fully custom rule sets.
#> rules redact_at block_at rate_guard selected
#> 1 14 0.4 0.75 FALSE FALSE
#> 2 14 0.4 0.75 FALSE FALSE
#> 3 18 0.3 0.60 FALSE FALSE
#> 4 17 0.4 0.75 TRUE FALSE
#> 5 16 0.4 0.75 FALSE FALSE
#> 6 8 0.8 0.95 FALSE FALSE
#> 7 23 0.4 0.70 TRUE TRUE
#> 8 0 0.4 0.75 FALSE FALSE
scan_prompt("hello", policy = "enterprise_default")
#> llmshieldr report
#> action: allow
#> risk_score: 0.000
#> findings: 0