prompt-sanitizer 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +7 -0
- data/CHANGELOG.md +34 -0
- data/README.md +269 -0
- data/lib/generators/prompt_sanitizer/install_generator.rb +38 -0
- data/lib/generators/prompt_sanitizer/templates/initializer.rb +36 -0
- data/lib/prompt_sanitizer/audit/base.rb +105 -0
- data/lib/prompt_sanitizer/audit/memory_audit_log.rb +86 -0
- data/lib/prompt_sanitizer/engines/ner_engine.rb +279 -0
- data/lib/prompt_sanitizer/engines/regex_engine.rb +216 -0
- data/lib/prompt_sanitizer/engines/secrets_engine.rb +230 -0
- data/lib/prompt_sanitizer/entities.rb +56 -0
- data/lib/prompt_sanitizer/integrations/action_controller.rb +64 -0
- data/lib/prompt_sanitizer/integrations/active_job.rb +79 -0
- data/lib/prompt_sanitizer/integrations/middleware.rb +153 -0
- data/lib/prompt_sanitizer/modes.rb +26 -0
- data/lib/prompt_sanitizer/railtie.rb +44 -0
- data/lib/prompt_sanitizer/result.rb +37 -0
- data/lib/prompt_sanitizer/sanitizer.rb +221 -0
- data/lib/prompt_sanitizer/session.rb +97 -0
- data/lib/prompt_sanitizer/synthetic.rb +152 -0
- data/lib/prompt_sanitizer/vault.rb +88 -0
- data/lib/prompt_sanitizer/version.rb +5 -0
- data/lib/prompt_sanitizer.rb +110 -0
- metadata +131 -0
|
@@ -0,0 +1,110 @@
|
|
|
1
|
+
# frozen_string_literal: true
|
|
2
|
+
|
|
3
|
+
require_relative "prompt_sanitizer/version"
|
|
4
|
+
|
|
5
|
+
module PromptSanitizer
|
|
6
|
+
class Error < StandardError; end
|
|
7
|
+
class ConfigurationError < Error; end
|
|
8
|
+
class ModeError < Error; end
|
|
9
|
+
|
|
10
|
+
# Raised when on_detect: :block and PII is found.
|
|
11
|
+
# @example
|
|
12
|
+
# rescue PromptSanitizer::PIIDetectedError => e
|
|
13
|
+
# e.entities # => Array<DetectedEntity>
|
|
14
|
+
class PIIDetectedError < Error
|
|
15
|
+
attr_reader :entities
|
|
16
|
+
|
|
17
|
+
def initialize(entities)
|
|
18
|
+
@entities = entities
|
|
19
|
+
types = entities.map { |e| e.entity_type }.uniq.join(", ")
|
|
20
|
+
super("PII detected and blocked (types: #{types})")
|
|
21
|
+
end
|
|
22
|
+
end
|
|
23
|
+
|
|
24
|
+
# ── Configuration ──────────────────────────────────────────────────────────
|
|
25
|
+
|
|
26
|
+
class Configuration
|
|
27
|
+
# Detection mode: :fast, :smart, :full
|
|
28
|
+
attr_accessor :mode
|
|
29
|
+
|
|
30
|
+
# NER backend (used in :smart / :full mode): :informers, :mitie
|
|
31
|
+
attr_accessor :ner_backend
|
|
32
|
+
|
|
33
|
+
# NER model variant: "distilbert" (66 MB, default) or "bert-base" (110 MB)
|
|
34
|
+
attr_accessor :ner_model
|
|
35
|
+
|
|
36
|
+
# Vault persistence: :memory (default), :rails_cache, :active_record
|
|
37
|
+
attr_accessor :vault_store
|
|
38
|
+
|
|
39
|
+
# Audit log backend: :none (default), :memory, :active_record
|
|
40
|
+
attr_accessor :audit_log
|
|
41
|
+
|
|
42
|
+
# Locale for synthetic replacements (Faker locale string, e.g. "en", "fr")
|
|
43
|
+
attr_accessor :locale
|
|
44
|
+
|
|
45
|
+
def initialize
|
|
46
|
+
@mode = :fast
|
|
47
|
+
@ner_backend = :informers
|
|
48
|
+
@ner_model = "distilbert"
|
|
49
|
+
@vault_store = :memory
|
|
50
|
+
@audit_log = :none
|
|
51
|
+
@locale = "en"
|
|
52
|
+
end
|
|
53
|
+
end
|
|
54
|
+
|
|
55
|
+
# ── Global helpers ──────────────────────────────────────────────────────────
|
|
56
|
+
|
|
57
|
+
class << self
|
|
58
|
+
def configuration
|
|
59
|
+
@configuration ||= Configuration.new
|
|
60
|
+
end
|
|
61
|
+
|
|
62
|
+
# PromptSanitizer.configure do |config|
|
|
63
|
+
# config.mode = :smart
|
|
64
|
+
# end
|
|
65
|
+
def configure
|
|
66
|
+
yield configuration
|
|
67
|
+
end
|
|
68
|
+
|
|
69
|
+
# Returns (and memoises) a Sanitizer built from the global configuration.
|
|
70
|
+
# Safe to call from multiple threads once initialised.
|
|
71
|
+
def sanitizer
|
|
72
|
+
@sanitizer ||= Sanitizer.new(
|
|
73
|
+
mode: configuration.mode,
|
|
74
|
+
ner_backend: configuration.ner_backend,
|
|
75
|
+
ner_model: configuration.ner_model,
|
|
76
|
+
audit_log: configuration.audit_log,
|
|
77
|
+
locale: configuration.locale
|
|
78
|
+
)
|
|
79
|
+
end
|
|
80
|
+
|
|
81
|
+
# Reset global state — intended for use in tests only.
|
|
82
|
+
def reset!
|
|
83
|
+
@configuration = nil
|
|
84
|
+
@sanitizer = nil
|
|
85
|
+
end
|
|
86
|
+
end
|
|
87
|
+
end
|
|
88
|
+
|
|
89
|
+
# Auto-require remaining library files (loaded after the module is defined
|
|
90
|
+
# so they can reference PromptSanitizer constants without forward-ref errors).
|
|
91
|
+
require_relative "prompt_sanitizer/entities"
|
|
92
|
+
require_relative "prompt_sanitizer/modes"
|
|
93
|
+
require_relative "prompt_sanitizer/result"
|
|
94
|
+
require_relative "prompt_sanitizer/vault"
|
|
95
|
+
require_relative "prompt_sanitizer/engines/regex_engine"
|
|
96
|
+
require_relative "prompt_sanitizer/engines/secrets_engine"
|
|
97
|
+
require_relative "prompt_sanitizer/engines/ner_engine"
|
|
98
|
+
require_relative "prompt_sanitizer/synthetic"
|
|
99
|
+
require_relative "prompt_sanitizer/audit/base"
|
|
100
|
+
require_relative "prompt_sanitizer/audit/memory_audit_log"
|
|
101
|
+
require_relative "prompt_sanitizer/session"
|
|
102
|
+
require_relative "prompt_sanitizer/sanitizer"
|
|
103
|
+
|
|
104
|
+
# Rails integrations — loaded only when Rails is present.
|
|
105
|
+
if defined?(Rails)
|
|
106
|
+
require_relative "prompt_sanitizer/railtie"
|
|
107
|
+
require_relative "prompt_sanitizer/integrations/middleware"
|
|
108
|
+
require_relative "prompt_sanitizer/integrations/action_controller"
|
|
109
|
+
require_relative "prompt_sanitizer/integrations/active_job"
|
|
110
|
+
end
|
metadata
ADDED
|
@@ -0,0 +1,131 @@
|
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
|
2
|
+
name: prompt-sanitizer
|
|
3
|
+
version: !ruby/object:Gem::Version
|
|
4
|
+
version: 0.1.0
|
|
5
|
+
platform: ruby
|
|
6
|
+
authors:
|
|
7
|
+
- Jeslor
|
|
8
|
+
autorequire:
|
|
9
|
+
bindir: bin
|
|
10
|
+
cert_chain: []
|
|
11
|
+
date: 2026-05-22 00:00:00.000000000 Z
|
|
12
|
+
dependencies:
|
|
13
|
+
- !ruby/object:Gem::Dependency
|
|
14
|
+
name: rspec
|
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
|
16
|
+
requirements:
|
|
17
|
+
- - "~>"
|
|
18
|
+
- !ruby/object:Gem::Version
|
|
19
|
+
version: '3.13'
|
|
20
|
+
type: :development
|
|
21
|
+
prerelease: false
|
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
23
|
+
requirements:
|
|
24
|
+
- - "~>"
|
|
25
|
+
- !ruby/object:Gem::Version
|
|
26
|
+
version: '3.13'
|
|
27
|
+
- !ruby/object:Gem::Dependency
|
|
28
|
+
name: rake
|
|
29
|
+
requirement: !ruby/object:Gem::Requirement
|
|
30
|
+
requirements:
|
|
31
|
+
- - "~>"
|
|
32
|
+
- !ruby/object:Gem::Version
|
|
33
|
+
version: '13.0'
|
|
34
|
+
type: :development
|
|
35
|
+
prerelease: false
|
|
36
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
37
|
+
requirements:
|
|
38
|
+
- - "~>"
|
|
39
|
+
- !ruby/object:Gem::Version
|
|
40
|
+
version: '13.0'
|
|
41
|
+
- !ruby/object:Gem::Dependency
|
|
42
|
+
name: rack
|
|
43
|
+
requirement: !ruby/object:Gem::Requirement
|
|
44
|
+
requirements:
|
|
45
|
+
- - "~>"
|
|
46
|
+
- !ruby/object:Gem::Version
|
|
47
|
+
version: '3.0'
|
|
48
|
+
type: :development
|
|
49
|
+
prerelease: false
|
|
50
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
51
|
+
requirements:
|
|
52
|
+
- - "~>"
|
|
53
|
+
- !ruby/object:Gem::Version
|
|
54
|
+
version: '3.0'
|
|
55
|
+
- !ruby/object:Gem::Dependency
|
|
56
|
+
name: faker
|
|
57
|
+
requirement: !ruby/object:Gem::Requirement
|
|
58
|
+
requirements:
|
|
59
|
+
- - "~>"
|
|
60
|
+
- !ruby/object:Gem::Version
|
|
61
|
+
version: '3.0'
|
|
62
|
+
type: :development
|
|
63
|
+
prerelease: false
|
|
64
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
65
|
+
requirements:
|
|
66
|
+
- - "~>"
|
|
67
|
+
- !ruby/object:Gem::Version
|
|
68
|
+
version: '3.0'
|
|
69
|
+
description: |
|
|
70
|
+
A lightweight, production-ready PII sanitizer built specifically for LLM
|
|
71
|
+
pipelines. Detects and redacts emails, phone numbers, SSNs, credit cards,
|
|
72
|
+
API keys, JWTs, AWS credentials, and more before prompts reach any model
|
|
73
|
+
API. Supports bidirectional sanitization (strip before, restore after) and
|
|
74
|
+
multi-turn session vaults. Zero cloud calls. GDPR & HIPAA ready.
|
|
75
|
+
email:
|
|
76
|
+
- hi@jeslor.com
|
|
77
|
+
executables: []
|
|
78
|
+
extensions: []
|
|
79
|
+
extra_rdoc_files: []
|
|
80
|
+
files:
|
|
81
|
+
- CHANGELOG.md
|
|
82
|
+
- README.md
|
|
83
|
+
- lib/generators/prompt_sanitizer/install_generator.rb
|
|
84
|
+
- lib/generators/prompt_sanitizer/templates/initializer.rb
|
|
85
|
+
- lib/prompt_sanitizer.rb
|
|
86
|
+
- lib/prompt_sanitizer/audit/base.rb
|
|
87
|
+
- lib/prompt_sanitizer/audit/memory_audit_log.rb
|
|
88
|
+
- lib/prompt_sanitizer/engines/ner_engine.rb
|
|
89
|
+
- lib/prompt_sanitizer/engines/regex_engine.rb
|
|
90
|
+
- lib/prompt_sanitizer/engines/secrets_engine.rb
|
|
91
|
+
- lib/prompt_sanitizer/entities.rb
|
|
92
|
+
- lib/prompt_sanitizer/integrations/action_controller.rb
|
|
93
|
+
- lib/prompt_sanitizer/integrations/active_job.rb
|
|
94
|
+
- lib/prompt_sanitizer/integrations/middleware.rb
|
|
95
|
+
- lib/prompt_sanitizer/modes.rb
|
|
96
|
+
- lib/prompt_sanitizer/railtie.rb
|
|
97
|
+
- lib/prompt_sanitizer/result.rb
|
|
98
|
+
- lib/prompt_sanitizer/sanitizer.rb
|
|
99
|
+
- lib/prompt_sanitizer/session.rb
|
|
100
|
+
- lib/prompt_sanitizer/synthetic.rb
|
|
101
|
+
- lib/prompt_sanitizer/vault.rb
|
|
102
|
+
- lib/prompt_sanitizer/version.rb
|
|
103
|
+
homepage: https://www.jeslor.com/prompt-sanitizer
|
|
104
|
+
licenses:
|
|
105
|
+
- MIT
|
|
106
|
+
metadata:
|
|
107
|
+
homepage_uri: https://www.jeslor.com/prompt-sanitizer
|
|
108
|
+
source_code_uri: https://github.com/jeslor/prompt-sanitizer/tree/main/packages/ruby
|
|
109
|
+
changelog_uri: https://github.com/jeslor/prompt-sanitizer/blob/main/packages/ruby/CHANGELOG.md
|
|
110
|
+
bug_tracker_uri: https://github.com/jeslor/prompt-sanitizer/issues
|
|
111
|
+
documentation_uri: https://rubydoc.info/gems/prompt-sanitizer
|
|
112
|
+
post_install_message:
|
|
113
|
+
rdoc_options: []
|
|
114
|
+
require_paths:
|
|
115
|
+
- lib
|
|
116
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
|
117
|
+
requirements:
|
|
118
|
+
- - ">="
|
|
119
|
+
- !ruby/object:Gem::Version
|
|
120
|
+
version: '3.1'
|
|
121
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
122
|
+
requirements:
|
|
123
|
+
- - ">="
|
|
124
|
+
- !ruby/object:Gem::Version
|
|
125
|
+
version: '0'
|
|
126
|
+
requirements: []
|
|
127
|
+
rubygems_version: 3.5.22
|
|
128
|
+
signing_key:
|
|
129
|
+
specification_version: 4
|
|
130
|
+
summary: Bidirectional PII sanitizer for LLM pipelines — Ruby edition
|
|
131
|
+
test_files: []
|