Safety Guardrails for Large Language Model Workflows
![[R logo]](figures/logo.png)
![[Up]](../../../doc/html/left.jpg)
![[Top]](../../../doc/html/up.jpg)
Documentation for package ‘llmshieldr’ version 0.1.0
Help Pages
| add_rule | Add a rule to a policy |
| available_policies | List available built-in policies |
| build_policy | Build a guardrail policy |
| builtin_rules | Built-in rule helpers |
| evaluate_security_cases | Evaluate scanner behavior on a labeled corpus |
| example_prompts | Example prompts |
| explain_findings | Explain findings |
| list_rules | List policy rules |
| ollama_reviewer | Create a local Ollama reviewer |
| policy | Load a built-in policy |
| policy_controls | Configure policy-level control actions |
| preflight_check | Preflight-check a prompt |
| print.shieldr_policy | Print a 'shieldr_policy' |
| print.shieldr_report | Print a 'shieldr_report' |
| rate_guard | Create or check a rate guard |
| redaction_strategy | Configure redaction behavior |
| remote_reviewer | Create a remote semantic reviewer |
| remove_rule | Remove a rule from a policy |
| reviewer_prompt | Return the default semantic reviewer prompt |
| rule_agency_language | Built-in rule helpers |
| rule_diagnosis_claim | Built-in rule helpers |
| rule_financial_advice | Built-in rule helpers |
| rule_injection_basic | Built-in rule helpers |
| rule_injection_indirect | Built-in rule helpers |
| rule_nlp_intent | Built-in rule helpers |
| rule_phi_condition | Built-in rule helpers |
| rule_pii_email | Built-in rule helpers |
| rule_pii_phone | Built-in rule helpers |
| rule_pii_ssn | Built-in rule helpers |
| rule_secrets_api_key | Built-in rule helpers |
| rule_secrets_aws | Built-in rule helpers |
| rule_secrets_bearer | Built-in rule helpers |
| rule_secrets_password | Built-in rule helpers |
| rule_system_prompt_leak | Built-in rule helpers |
| scanner_options | Configure optional text scanners |
| scan_context | Scan retrieved context rows |
| scan_conversation | Scan a conversation while preserving message roles |
| scan_output | Scan model output |
| scan_prompt | Scan a prompt |
| scan_stream | Scan streamed output chunks with rolling context |
| scan_tool_call | Scan a tool call before execution |
| scan_tool_output | Scan tool output before it re-enters model context |
| secure_chat | Run a guarded chat call |
| shieldr_audit | Construct a 'shieldr_audit' |
| shieldr_policy | Construct a 'shieldr_policy' |
| shieldr_report | Construct a 'shieldr_report' |
| shieldr_result | Construct a 'shieldr_result' |
| shieldr_rule | Construct a 'shieldr_rule' |
| shield_ollama | Guard an Ollama chat workflow |
| trust_boundary | Wrap a chat object in a trust boundary |
| write_audit_log | Write an audit log |