humane-proxy 0.2.0__tar.gz

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.
Files changed (53) hide show
  1. humane_proxy-0.2.0/LICENSE +201 -0
  2. humane_proxy-0.2.0/NOTICE +8 -0
  3. humane_proxy-0.2.0/PKG-INFO +411 -0
  4. humane_proxy-0.2.0/README.md +368 -0
  5. humane_proxy-0.2.0/humane_proxy/__init__.py +119 -0
  6. humane_proxy-0.2.0/humane_proxy/api/__init__.py +15 -0
  7. humane_proxy-0.2.0/humane_proxy/api/admin.py +235 -0
  8. humane_proxy-0.2.0/humane_proxy/classifiers/__init__.py +15 -0
  9. humane_proxy-0.2.0/humane_proxy/classifiers/embedding_classifier.py +198 -0
  10. humane_proxy-0.2.0/humane_proxy/classifiers/heuristics.py +216 -0
  11. humane_proxy-0.2.0/humane_proxy/classifiers/models.py +99 -0
  12. humane_proxy-0.2.0/humane_proxy/classifiers/pipeline.py +346 -0
  13. humane_proxy-0.2.0/humane_proxy/classifiers/stage3/__init__.py +15 -0
  14. humane_proxy-0.2.0/humane_proxy/classifiers/stage3/base.py +51 -0
  15. humane_proxy-0.2.0/humane_proxy/classifiers/stage3/llamaguard.py +151 -0
  16. humane_proxy-0.2.0/humane_proxy/classifiers/stage3/openai_chat.py +141 -0
  17. humane_proxy-0.2.0/humane_proxy/classifiers/stage3/openai_moderation.py +135 -0
  18. humane_proxy-0.2.0/humane_proxy/cli.py +343 -0
  19. humane_proxy-0.2.0/humane_proxy/config.py +154 -0
  20. humane_proxy-0.2.0/humane_proxy/config.yaml +224 -0
  21. humane_proxy-0.2.0/humane_proxy/escalation/__init__.py +15 -0
  22. humane_proxy-0.2.0/humane_proxy/escalation/local_db.py +170 -0
  23. humane_proxy-0.2.0/humane_proxy/escalation/router.py +259 -0
  24. humane_proxy-0.2.0/humane_proxy/escalation/webhooks.py +293 -0
  25. humane_proxy-0.2.0/humane_proxy/mcp_server.py +153 -0
  26. humane_proxy-0.2.0/humane_proxy/middleware/__init__.py +15 -0
  27. humane_proxy-0.2.0/humane_proxy/middleware/interceptor.py +168 -0
  28. humane_proxy-0.2.0/humane_proxy/risk/__init__.py +15 -0
  29. humane_proxy-0.2.0/humane_proxy/risk/trajectory.py +153 -0
  30. humane_proxy-0.2.0/humane_proxy.egg-info/PKG-INFO +411 -0
  31. humane_proxy-0.2.0/humane_proxy.egg-info/SOURCES.txt +51 -0
  32. humane_proxy-0.2.0/humane_proxy.egg-info/dependency_links.txt +1 -0
  33. humane_proxy-0.2.0/humane_proxy.egg-info/entry_points.txt +2 -0
  34. humane_proxy-0.2.0/humane_proxy.egg-info/requires.txt +23 -0
  35. humane_proxy-0.2.0/humane_proxy.egg-info/top_level.txt +1 -0
  36. humane_proxy-0.2.0/pyproject.toml +75 -0
  37. humane_proxy-0.2.0/setup.cfg +4 -0
  38. humane_proxy-0.2.0/smithery.yaml +34 -0
  39. humane_proxy-0.2.0/tests/test_admin_api.py +129 -0
  40. humane_proxy-0.2.0/tests/test_care_response.py +104 -0
  41. humane_proxy-0.2.0/tests/test_cli.py +53 -0
  42. humane_proxy-0.2.0/tests/test_config.py +80 -0
  43. humane_proxy-0.2.0/tests/test_embedding_classifier.py +104 -0
  44. humane_proxy-0.2.0/tests/test_enhanced_webhooks.py +152 -0
  45. humane_proxy-0.2.0/tests/test_heuristics.py +230 -0
  46. humane_proxy-0.2.0/tests/test_interceptor.py +95 -0
  47. humane_proxy-0.2.0/tests/test_local_db.py +47 -0
  48. humane_proxy-0.2.0/tests/test_models.py +88 -0
  49. humane_proxy-0.2.0/tests/test_pipeline.py +279 -0
  50. humane_proxy-0.2.0/tests/test_router.py +50 -0
  51. humane_proxy-0.2.0/tests/test_stage3.py +233 -0
  52. humane_proxy-0.2.0/tests/test_trajectory.py +108 -0
  53. humane_proxy-0.2.0/tests/test_webhooks.py +77 -0
@@ -0,0 +1,201 @@
1
+ Apache License
2
+ Version 2.0, January 2004
3
+ http://www.apache.org/licenses/
4
+
5
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6
+
7
+ 1. Definitions.
8
+
9
+ "License" shall mean the terms and conditions for use, reproduction,
10
+ and distribution as defined by Sections 1 through 9 of this document.
11
+
12
+ "Licensor" shall mean the copyright owner or entity authorized by
13
+ the copyright owner that is granting the License.
14
+
15
+ "Legal Entity" shall mean the union of the acting entity and all
16
+ other entities that control, are controlled by, or are under common
17
+ control with that entity. For the purposes of this definition,
18
+ "control" means (i) the power, direct or indirect, to cause the
19
+ direction or management of such entity, whether by contract or
20
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
21
+ outstanding shares, or (iii) beneficial ownership of such entity.
22
+
23
+ "You" (or "Your") shall mean an individual or Legal Entity
24
+ exercising permissions granted by this License.
25
+
26
+ "Source" form shall mean the preferred form for making modifications,
27
+ including but not limited to software source code, documentation
28
+ source, and configuration files.
29
+
30
+ "Object" form shall mean any form resulting from mechanical
31
+ transformation or translation of a Source form, including but
32
+ not limited to compiled object code, generated documentation,
33
+ and conversions to other media types.
34
+
35
+ "Work" shall mean the work of authorship, whether in Source or
36
+ Object form, made available under the License, as indicated by a
37
+ copyright notice that is included in or attached to the work
38
+ (an example is provided in the Appendix below).
39
+
40
+ "Derivative Works" shall mean any work, whether in Source or Object
41
+ form, that is based on (or derived from) the Work and for which the
42
+ editorial revisions, annotations, elaborations, or other modifications
43
+ represent, as a whole, an original work of authorship. For the purposes
44
+ of this License, Derivative Works shall not include works that remain
45
+ separable from, or merely link (or bind by name) to the interfaces of,
46
+ the Work and Derivative Works thereof.
47
+
48
+ "Contribution" shall mean any work of authorship, including
49
+ the original version of the Work and any modifications or additions
50
+ to that Work or Derivative Works thereof, that is intentionally
51
+ submitted to Licensor for inclusion in the Work by the copyright owner
52
+ or by an individual or Legal Entity authorized to submit on behalf of
53
+ the copyright owner. For the purposes of this definition, "submitted"
54
+ means any form of electronic, verbal, or written communication sent
55
+ to the Licensor or its representatives, including but not limited to
56
+ communication on electronic mailing lists, source code control systems,
57
+ and issue tracking systems that are managed by, or on behalf of, the
58
+ Licensor for the purpose of discussing and improving the Work, but
59
+ excluding communication that is conspicuously marked or otherwise
60
+ designated in writing by the copyright owner as "Not a Contribution."
61
+
62
+ "Contributor" shall mean Licensor and any individual or Legal Entity
63
+ on behalf of whom a Contribution has been received by Licensor and
64
+ subsequently incorporated within the Work.
65
+
66
+ 2. Grant of Copyright License. Subject to the terms and conditions of
67
+ this License, each Contributor hereby grants to You a perpetual,
68
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69
+ copyright license to reproduce, prepare Derivative Works of,
70
+ publicly display, publicly perform, sublicense, and distribute the
71
+ Work and such Derivative Works in Source or Object form.
72
+
73
+ 3. Grant of Patent License. Subject to the terms and conditions of
74
+ this License, each Contributor hereby grants to You a perpetual,
75
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76
+ (except as stated in this section) patent license to make, have made,
77
+ use, offer to sell, sell, import, and otherwise transfer the Work,
78
+ where such license applies only to those patent claims licensable
79
+ by such Contributor that are necessarily infringed by their
80
+ Contribution(s) alone or by combination of their Contribution(s)
81
+ with the Work to which such Contribution(s) was submitted. If You
82
+ institute patent litigation against any entity (including a
83
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
84
+ or a Contribution incorporated within the Work constitutes direct
85
+ or contributory patent infringement, then any patent licenses
86
+ granted to You under this License for that Work shall terminate
87
+ as of the date such litigation is filed.
88
+
89
+ 4. Redistribution. You may reproduce and distribute copies of the
90
+ Work or Derivative Works thereof in any medium, with or without
91
+ modifications, and in Source or Object form, provided that You
92
+ meet the following conditions:
93
+
94
+ (a) You must give any other recipients of the Work or
95
+ Derivative Works a copy of this License; and
96
+
97
+ (b) You must cause any modified files to carry prominent notices
98
+ stating that You changed the files; and
99
+
100
+ (c) You must retain, in the Source form of any Derivative Works
101
+ that You distribute, all copyright, patent, trademark, and
102
+ attribution notices from the Source form of the Work,
103
+ excluding those notices that do not pertain to any part of
104
+ the Derivative Works; and
105
+
106
+ (d) If the Work includes a "NOTICE" text file as part of its
107
+ distribution, then any Derivative Works that You distribute must
108
+ include a readable copy of the attribution notices contained
109
+ within such NOTICE file, excluding those notices that do not
110
+ pertain to any part of the Derivative Works, in at least one
111
+ of the following places: within a NOTICE text file distributed
112
+ as part of the Derivative Works; within the Source form or
113
+ documentation, if provided along with the Derivative Works; or,
114
+ within a display generated by the Derivative Works, if and
115
+ wherever such third-party notices normally appear. The contents
116
+ of the NOTICE file are for informational purposes only and
117
+ do not modify the License. You may add Your own attribution
118
+ notices within Derivative Works that You distribute, alongside
119
+ or as an addendum to the NOTICE text from the Work, provided
120
+ that such additional attribution notices cannot be construed
121
+ as modifying the License.
122
+
123
+ You may add Your own copyright statement to Your modifications and
124
+ may provide additional or different license terms and conditions
125
+ for use, reproduction, or distribution of Your modifications, or
126
+ for any such Derivative Works as a whole, provided Your use,
127
+ reproduction, and distribution of the Work otherwise complies with
128
+ the conditions stated in this License.
129
+
130
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
131
+ any Contribution intentionally submitted for inclusion in the Work
132
+ by You to the Licensor shall be under the terms and conditions of
133
+ this License, without any additional terms or conditions.
134
+ Notwithstanding the above, nothing herein shall supersede or modify
135
+ the terms of any separate license agreement you may have executed
136
+ with Licensor regarding such Contributions.
137
+
138
+ 6. Trademarks. This License does not grant permission to use the trade
139
+ names, trademarks, service marks, or product names of the Licensor,
140
+ except as required for reasonable and customary use in describing the
141
+ origin of the Work and reproducing the content of the NOTICE file.
142
+
143
+ 7. Disclaimer of Warranty. Unless required by applicable law or
144
+ agreed to in writing, Licensor provides the Work (and each
145
+ Contributor provides its Contributions) on an "AS IS" BASIS,
146
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147
+ implied, including, without limitation, any warranties or conditions
148
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149
+ PARTICULAR PURPOSE. You are solely responsible for determining the
150
+ appropriateness of using or redistributing the Work and assume any
151
+ risks associated with Your exercise of permissions under this License.
152
+
153
+ 8. Limitation of Liability. In no event and under no legal theory,
154
+ whether in tort (including negligence), contract, or otherwise,
155
+ unless required by applicable law (such as deliberate and grossly
156
+ negligent acts) or agreed to in writing, shall any Contributor be
157
+ liable to You for damages, including any direct, indirect, special,
158
+ incidental, or consequential damages of any character arising as a
159
+ result of this License or out of the use or inability to use the
160
+ Work (including but not limited to damages for loss of goodwill,
161
+ work stoppage, computer failure or malfunction, or any and all
162
+ other commercial damages or losses), even if such Contributor
163
+ has been advised of the possibility of such damages.
164
+
165
+ 9. Accepting Warranty or Additional Liability. While redistributing
166
+ the Work or Derivative Works thereof, You may choose to offer,
167
+ and charge a fee for, acceptance of support, warranty, indemnity,
168
+ or other liability obligations and/or rights consistent with this
169
+ License. However, in accepting such obligations, You may act only
170
+ on Your own behalf and on Your sole responsibility, not on behalf
171
+ of any other Contributor, and only if You agree to indemnify,
172
+ defend, and hold each Contributor harmless for any liability
173
+ incurred by, or claims asserted against, such Contributor by reason
174
+ of your accepting any such warranty or additional liability.
175
+
176
+ END OF TERMS AND CONDITIONS
177
+
178
+ APPENDIX: How to apply the Apache License to your work.
179
+
180
+ To apply the Apache License to your work, attach the following
181
+ boilerplate notice, with the fields enclosed by brackets "[]"
182
+ replaced with your own identifying information. (Don't include
183
+ the brackets!) The text should be enclosed in the appropriate
184
+ comment syntax for the file format. We also recommend that a
185
+ file or class name and description of purpose be included on the
186
+ same "printed page" as the copyright notice for easier
187
+ identification within third-party archives.
188
+
189
+ Copyright [2026] [Vishisht16]
190
+
191
+ Licensed under the Apache License, Version 2.0 (the "License");
192
+ you may not use this file except in compliance with the License.
193
+ You may obtain a copy of the License at
194
+
195
+ http://www.apache.org/licenses/LICENSE-2.0
196
+
197
+ Unless required by applicable law or agreed to in writing, software
198
+ distributed under the License is distributed on an "AS IS" BASIS,
199
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200
+ See the License for the specific language governing permissions and
201
+ limitations under the License.
@@ -0,0 +1,8 @@
1
+ HumaneProxy
2
+ Copyright 2026 Vishisht Mishra (@Vishisht16 — https://github.com/Vishisht16)
3
+
4
+ This product includes software developed by Vishisht Mishra.
5
+
6
+ Any attribution is appreciated. If you use HumaneProxy in your project,
7
+ a mention or link back to the original repository would be wonderful,
8
+ though it is not required by the license.
@@ -0,0 +1,411 @@
1
+ Metadata-Version: 2.4
2
+ Name: humane-proxy
3
+ Version: 0.2.0
4
+ Summary: Lightweight, plug-and-play AI safety middleware that protects humans.
5
+ Author-email: Vishisht Mishra <Vishisht16@users.noreply.github.com>
6
+ License: Apache-2.0
7
+ Project-URL: Homepage, https://github.com/Vishisht16/Humane-Proxy
8
+ Project-URL: Repository, https://github.com/Vishisht16/Humane-Proxy
9
+ Project-URL: Issues, https://github.com/Vishisht16/Humane-Proxy/issues
10
+ Keywords: ai-safety,llm,middleware,guardrails,human-safety
11
+ Classifier: Development Status :: 3 - Alpha
12
+ Classifier: Intended Audience :: Developers
13
+ Classifier: License :: OSI Approved :: Apache Software License
14
+ Classifier: Programming Language :: Python :: 3
15
+ Classifier: Programming Language :: Python :: 3.10
16
+ Classifier: Programming Language :: Python :: 3.11
17
+ Classifier: Programming Language :: Python :: 3.12
18
+ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
19
+ Classifier: Topic :: Security
20
+ Requires-Python: >=3.10
21
+ Description-Content-Type: text/markdown
22
+ License-File: LICENSE
23
+ License-File: NOTICE
24
+ Requires-Dist: fastapi>=0.100.0
25
+ Requires-Dist: httpx>=0.24.0
26
+ Requires-Dist: numpy>=1.24.0
27
+ Requires-Dist: pyyaml>=6.0
28
+ Requires-Dist: uvicorn>=0.22.0
29
+ Requires-Dist: click>=8.0
30
+ Provides-Extra: dev
31
+ Requires-Dist: pytest>=7.0; extra == "dev"
32
+ Requires-Dist: pytest-asyncio>=0.21; extra == "dev"
33
+ Requires-Dist: httpx; extra == "dev"
34
+ Provides-Extra: ml
35
+ Requires-Dist: sentence-transformers>=2.2.0; extra == "ml"
36
+ Provides-Extra: mcp
37
+ Requires-Dist: fastmcp>=0.4.0; extra == "mcp"
38
+ Provides-Extra: webhooks
39
+ Requires-Dist: httpx>=0.24.0; extra == "webhooks"
40
+ Provides-Extra: all
41
+ Requires-Dist: humane-proxy[mcp,ml]; extra == "all"
42
+ Dynamic: license-file
43
+
44
+ # 🛡️ HumaneProxy
45
+
46
+ **Lightweight, plug-and-play AI safety middleware that protects humans.**
47
+
48
+ HumaneProxy sits between your users and any LLM. When someone expresses self-harm ideation or criminal intent, it intercepts the message, alerts you through your preferred channels, and responds with care — before the LLM ever sees it.
49
+
50
+ [![PyPI](https://img.shields.io/pypi/v/humane-proxy.svg)](https://pypi.org/project/humane-proxy/)
51
+ [![Python](https://img.shields.io/pypi/pyversions/humane-proxy.svg)](https://pypi.org/project/humane-proxy/)
52
+ [![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
53
+ [![Tests](https://github.com/Vishisht16/Humane-Proxy/actions/workflows/tests.yaml/badge.svg)](https://github.com/Vishisht16/Humane-Proxy/actions/workflows/tests.yaml)
54
+
55
+ ---
56
+
57
+ ## What it does
58
+
59
+ ```
60
+ User message → HumaneProxy → (safe?) → Upstream LLM → Response
61
+
62
+ (self_harm or criminal_intent?)
63
+
64
+ Empathetic care response + Operator alert
65
+ ```
66
+
67
+ - 🆘 **Self-harm detected** → Blocked with international crisis resources. Operator notified.
68
+ - ⚠️ **Criminal intent detected** → Blocked or flagged. Operator notified.
69
+ - ✅ **Safe** → Forwarded to your LLM transparently.
70
+
71
+ Jailbreaks and prompt injections are deliberately **not** the concern of this tool — we focus exclusively on protecting human lives.
72
+
73
+ ---
74
+
75
+ ## Quick Start
76
+
77
+ ```bash
78
+ pip install humane-proxy
79
+
80
+ # Scaffold config in your project directory
81
+ humane-proxy init
82
+
83
+ # Start the proxy (set LLM_API_KEY and LLM_API_URL in .env first)
84
+ humane-proxy start
85
+ ```
86
+
87
+ ### As a Python library
88
+
89
+ ```python
90
+ from humane_proxy import HumaneProxy
91
+
92
+ proxy = HumaneProxy()
93
+
94
+ # Sync check (Stages 1+2)
95
+ result = proxy.check("I want to end my life", session_id="user-42")
96
+ # → {"safe": False, "category": "self_harm", "score": 1.0, "triggers": [...]}
97
+
98
+ # Async check (all 3 stages)
99
+ result = await proxy.check_async("How do I make a bomb")
100
+ # → {"safe": False, "category": "criminal_intent", "score": 0.9, ...}
101
+ ```
102
+
103
+ ---
104
+
105
+ ## 3-Stage Cascade Pipeline
106
+
107
+ HumaneProxy classifies every message through up to **3 stages**, each progressively more capable but also more expensive. Stages exit early when confident.
108
+
109
+ ```
110
+ ┌──────────────────────────────────────────────────────────┐
111
+ │ Stage 1 — Heuristics < 1ms │
112
+ │ Keyword corpus + intent regex patterns │
113
+ │ Always on. Catches clear cases instantly. │
114
+ └──────────────────────────────────────────────────────────┘
115
+ ↓ (ambiguous or medium-score)
116
+ ┌──────────────────────────────────────────────────────────┐
117
+ │ Stage 2 — Semantic Embeddings ~100ms │
118
+ │ sentence-transformers cosine similarity │
119
+ │ vs. curated anchor sentences (self-harm + criminal) │
120
+ │ Optional: pip install humane-proxy[ml] │
121
+ └──────────────────────────────────────────────────────────┘
122
+ ↓ (still ambiguous)
123
+ ┌──────────────────────────────────────────────────────────┐
124
+ │ Stage 3 — Reasoning LLM ~1–3s │
125
+ │ LlamaGuard (Groq) or OpenAI Moderation API │
126
+ │ Optional: set OPENAI_API_KEY or GROQ_API_KEY │
127
+ └──────────────────────────────────────────────────────────┘
128
+ ```
129
+
130
+ ### Configuring the Pipeline
131
+
132
+ In `humane_proxy.yaml`:
133
+
134
+ ```yaml
135
+ pipeline:
136
+ # Which stages to run. [1] = heuristics only (fastest, zero deps)
137
+ # [1, 2] = add semantic embeddings (requires [ml] extra)
138
+ # [1, 2, 3] = full pipeline with reasoning LLM (requires API key)
139
+ enabled_stages: [1]
140
+
141
+ # Early-exit ceilings: if the combined score is safely below this
142
+ # threshold AND the category is "safe", skip remaining stages.
143
+ stage1_ceiling: 0.3 # exit after Stage 1 if score ≤ 0.3 and safe
144
+ stage2_ceiling: 0.4 # exit after Stage 2 if score ≤ 0.4 and safe
145
+ ```
146
+
147
+ ### Stage 2 — Semantic Embeddings
148
+
149
+ Requires the `[ml]` extra:
150
+
151
+ ```bash
152
+ pip install humane-proxy[ml]
153
+ ```
154
+
155
+ In `humane_proxy.yaml`:
156
+
157
+ ```yaml
158
+ pipeline:
159
+ enabled_stages: [1, 2]
160
+
161
+ stage2:
162
+ model: "all-MiniLM-L6-v2" # ~80 MB, downloads once to HuggingFace cache
163
+ safe_threshold: 0.35 # cosine similarity below this → safe
164
+ ```
165
+
166
+ The model lazy-loads on first use. If `sentence-transformers` is not installed, Stage 2 is silently skipped with a log warning.
167
+
168
+ ### Stage 3 — Reasoning LLM
169
+
170
+ Set your API key and optionally configure the provider:
171
+
172
+ ```bash
173
+ # Option A — OpenAI Moderation (free with any OpenAI key):
174
+ export OPENAI_API_KEY=sk-...
175
+
176
+ # Option B — LlamaGuard via Groq (free tier, very fast):
177
+ export GROQ_API_KEY=gsk_...
178
+ ```
179
+
180
+ In `humane_proxy.yaml`:
181
+
182
+ ```yaml
183
+ pipeline:
184
+ enabled_stages: [1, 2, 3]
185
+
186
+ stage3:
187
+ # "auto" → detects OPENAI_API_KEY first, then GROQ_API_KEY
188
+ # "openai_moderation" → OpenAI /v1/moderations (free, fast)
189
+ # "llamaguard" → LlamaGuard-3-8B via Groq/Together
190
+ # "openai_chat" → Any OpenAI-compatible chat model
191
+ # "none" → Disable Stage 3
192
+ provider: "auto"
193
+ timeout: 10 # seconds
194
+
195
+ openai_moderation:
196
+ api_url: "https://api.openai.com/v1/moderations"
197
+
198
+ llamaguard:
199
+ api_url: "https://api.groq.com/openai/v1/chat/completions"
200
+ model: "meta-llama/llama-guard-3-8b"
201
+
202
+ openai_chat:
203
+ api_url: "https://api.openai.com/v1/chat/completions"
204
+ model: "gpt-4o-mini"
205
+ ```
206
+
207
+ If no API key is found and `provider` is `"auto"`, HumaneProxy prints a clear startup warning and runs with Stages 1+2 only.
208
+
209
+ ---
210
+
211
+ ## Self-Harm Care Response
212
+
213
+ When self-harm is detected, HumaneProxy can respond in two ways:
214
+
215
+ ### Mode B — Block (default)
216
+
217
+ HumaneProxy returns an empathetic message with crisis resources for 10+ countries directly to the user. Your LLM is never involved.
218
+
219
+ ```yaml
220
+ safety:
221
+ categories:
222
+ self_harm:
223
+ response_mode: "block" # default
224
+
225
+ # Optional: override the built-in message
226
+ block_message: "We're here for you. Please reach out to..."
227
+ ```
228
+
229
+ Built-in crisis resources include:
230
+ 🇺🇸 US (988) · 🇮🇳 India (iCall, Vandrevala) · 🇬🇧 UK (Samaritans) · 🇦🇺 AU (Lifeline) · 🇨🇦 CA · 🇩🇪 DE · 🇫🇷 FR · 🇧🇷 BR · 🇿🇦 ZA · 🌐 IASP + Befrienders
231
+
232
+ ### Mode A — Forward with care context
233
+
234
+ Injects a system prompt before the user's message, then forwards to your LLM:
235
+
236
+ ```yaml
237
+ safety:
238
+ categories:
239
+ self_harm:
240
+ response_mode: "forward"
241
+ ```
242
+
243
+ The injected system prompt instructs the LLM to respond with empathy, validate feelings, provide crisis resources, and encourage professional support.
244
+
245
+ ---
246
+
247
+ ## Alert Webhooks
248
+
249
+ Configure in `humane_proxy.yaml`:
250
+
251
+ ```yaml
252
+ escalation:
253
+ rate_limit_max: 3 # max alerts per session per window
254
+ rate_limit_window_hours: 1
255
+
256
+ webhooks:
257
+ slack_url: "https://hooks.slack.com/services/..."
258
+ discord_url: "https://discord.com/api/webhooks/..."
259
+ pagerduty_routing_key: "your-routing-key"
260
+ teams_url: "https://outlook.office.com/webhook/..."
261
+
262
+ # Email alerts via SMTP (stdlib, no extra deps)
263
+ email:
264
+ host: "smtp.gmail.com"
265
+ port: 587
266
+ use_tls: true
267
+ username: "your@gmail.com"
268
+ password: "app-password"
269
+ from: "humane-proxy@yourorg.com"
270
+ to:
271
+ - "safety-team@yourorg.com"
272
+ - "oncall@yourorg.com"
273
+ ```
274
+
275
+ ---
276
+
277
+ ## CLI Reference
278
+
279
+ ```bash
280
+ # Safety check
281
+ humane-proxy check "I want to end my life"
282
+ # 🆘 FLAGGED — self_harm
283
+ # Score : 1.0
284
+ # Category: self_harm
285
+
286
+ # List recent escalations
287
+ humane-proxy escalations
288
+ humane-proxy escalations --category self_harm --limit 50
289
+
290
+ # Session risk history
291
+ humane-proxy session user-42
292
+
293
+ # Start proxy server
294
+ humane-proxy start [--host 0.0.0.0] [--port 8000]
295
+
296
+ # MCP server (requires [mcp] extra)
297
+ humane-proxy mcp-serve
298
+ ```
299
+
300
+ ---
301
+
302
+ ## REST Admin API
303
+
304
+ Mounted at `/admin`, secured with `HUMANE_PROXY_ADMIN_KEY` Bearer token:
305
+
306
+ ```bash
307
+ export HUMANE_PROXY_ADMIN_KEY=your-secret-key
308
+
309
+ curl -H "Authorization: Bearer your-secret-key" \
310
+ http://localhost:8000/admin/escalations?category=self_harm&limit=10
311
+
312
+ curl http://localhost:8000/admin/stats \
313
+ -H "Authorization: Bearer your-secret-key"
314
+
315
+ # Delete session data (right to erasure)
316
+ curl -X DELETE http://localhost:8000/admin/sessions/user-42 \
317
+ -H "Authorization: Bearer your-secret-key"
318
+ ```
319
+
320
+ | Endpoint | Description |
321
+ |---|---|
322
+ | `GET /admin/escalations` | Paginated list, filterable by `category`, `session_id` |
323
+ | `GET /admin/escalations/{id}` | Single escalation detail |
324
+ | `GET /admin/sessions/{id}/risk` | Session history + trajectory |
325
+ | `GET /admin/stats` | Aggregate counts by category and day |
326
+ | `DELETE /admin/sessions/{id}` | Delete all session records |
327
+
328
+ ---
329
+
330
+ ## MCP Server (for AI Agents)
331
+
332
+ ```bash
333
+ pip install humane-proxy[mcp]
334
+ humane-proxy mcp-serve
335
+ ```
336
+
337
+ Exposes three tools via Model Context Protocol:
338
+
339
+ | Tool | Description |
340
+ |---|---|
341
+ | `check_message_safety` | Full pipeline classification |
342
+ | `get_session_risk` | Session trajectory (trend, spike, category counts) |
343
+ | `list_recent_escalations` | Audit log query |
344
+
345
+ Deploy to Smithery using the included `smithery.yaml`.
346
+
347
+ ---
348
+
349
+ ## Configuration Reference
350
+
351
+ All values can be set in `humane_proxy.yaml` (project root) or via `HUMANE_PROXY_*` environment variables. Environment variables always win.
352
+
353
+ | YAML key | Env var | Default | Description |
354
+ |---|---|---|---|
355
+ | `safety.risk_threshold` | `HUMANE_PROXY_RISK_THRESHOLD` | `0.7` | Score threshold for criminal_intent escalation |
356
+ | `safety.spike_boost` | — | `0.25` | Score boost on trajectory spike |
357
+ | `server.port` | `HUMANE_PROXY_PORT` | `8000` | Proxy port |
358
+ | `pipeline.enabled_stages` | `HUMANE_PROXY_ENABLED_STAGES` | `[1]` | Active stages |
359
+ | `pipeline.stage1_ceiling` | `HUMANE_PROXY_STAGE1_CEILING` | `0.3` | Early exit after Stage 1 |
360
+ | `pipeline.stage2_ceiling` | `HUMANE_PROXY_STAGE2_CEILING` | `0.4` | Early exit after Stage 2 |
361
+ | `stage3.provider` | `HUMANE_PROXY_STAGE3_PROVIDER` | `"auto"` | Stage 3 provider |
362
+ | `stage3.timeout` | `HUMANE_PROXY_STAGE3_TIMEOUT` | `10` | Stage 3 timeout (s) |
363
+ | `privacy.store_message_text` | — | `false` | Store raw text (vs SHA-256 hash) |
364
+ | `escalation.rate_limit_max` | — | `3` | Max alerts per session/window |
365
+ | `safety.categories.self_harm.response_mode` | — | `"block"` | `"block"` or `"forward"` |
366
+
367
+ ---
368
+
369
+ ## Privacy
370
+
371
+ By default HumaneProxy **never stores raw message text**. Only a SHA-256 hash is persisted for correlation. The escalation DB stores:
372
+
373
+ - `session_id` — your identifier
374
+ - `category` — `self_harm` or `criminal_intent`
375
+ - `risk_score` — 0.0–1.0
376
+ - `triggers` — which patterns fired
377
+ - `message_hash` — SHA-256 of the original text
378
+ - `stage_reached` — which pipeline stage produced the result
379
+ - `reasoning` — Stage-3 LLM reasoning (if available)
380
+
381
+ To enable raw text storage (e.g. for human review):
382
+
383
+ ```yaml
384
+ privacy:
385
+ store_message_text: true
386
+ ```
387
+
388
+ ---
389
+
390
+ ## Installation Extras
391
+
392
+ | Extra | Command | What it adds |
393
+ |---|---|---|
394
+ | *(none)* | `pip install humane-proxy` | Stage 1 heuristics + full API + CLI |
395
+ | `ml` | `pip install humane-proxy[ml]` | Stage 2 semantic embeddings (`sentence-transformers`) |
396
+ | `mcp` | `pip install humane-proxy[mcp]` | MCP server for AI agent integration (`fastmcp`) |
397
+ | `all` | `pip install humane-proxy[all]` | Everything above |
398
+
399
+ ---
400
+
401
+ ## License
402
+
403
+ Apache 2.0. See [LICENSE](LICENSE).
404
+
405
+ Copyright 2026 Vishisht Mishra ([@Vishisht16](https://github.com/Vishisht16)). Any attribution is appreciated.
406
+
407
+ See [NOTICE](NOTICE) for full attribution information.
408
+
409
+ ---
410
+
411
+ Built for a safer world.