mimvo 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.
- mimvo-0.2.0/.gitignore +50 -0
- mimvo-0.2.0/CHANGELOG.md +258 -0
- mimvo-0.2.0/LICENSE +201 -0
- mimvo-0.2.0/PKG-INFO +684 -0
- mimvo-0.2.0/README.md +606 -0
- mimvo-0.2.0/benchmarks/README.md +70 -0
- mimvo-0.2.0/mimvo/__init__.py +53 -0
- mimvo-0.2.0/mimvo/_coerce.py +72 -0
- mimvo-0.2.0/mimvo/_version.py +11 -0
- mimvo-0.2.0/mimvo/checks/__init__.py +54 -0
- mimvo-0.2.0/mimvo/checks/base.py +149 -0
- mimvo-0.2.0/mimvo/checks/security/__init__.py +43 -0
- mimvo-0.2.0/mimvo/checks/security/_hf.py +439 -0
- mimvo-0.2.0/mimvo/checks/security/_http.py +62 -0
- mimvo-0.2.0/mimvo/checks/security/_phrases.py +180 -0
- mimvo-0.2.0/mimvo/checks/security/base.py +780 -0
- mimvo-0.2.0/mimvo/checks/security/injection/__init__.py +48 -0
- mimvo-0.2.0/mimvo/checks/security/injection/check.py +45 -0
- mimvo-0.2.0/mimvo/checks/security/injection/datasentinel.py +81 -0
- mimvo-0.2.0/mimvo/checks/security/injection/deepset_deberta.py +31 -0
- mimvo-0.2.0/mimvo/checks/security/injection/embedding_clf.py +76 -0
- mimvo-0.2.0/mimvo/checks/security/injection/heuristic.py +429 -0
- mimvo-0.2.0/mimvo/checks/security/injection/known_answer.py +76 -0
- mimvo-0.2.0/mimvo/checks/security/injection/lakera_guard.py +129 -0
- mimvo-0.2.0/mimvo/checks/security/injection/probes.py +80 -0
- mimvo-0.2.0/mimvo/checks/security/injection/prompt_guard.py +37 -0
- mimvo-0.2.0/mimvo/checks/security/injection/prompt_shield.py +127 -0
- mimvo-0.2.0/mimvo/checks/security/injection/protectai_deberta.py +31 -0
- mimvo-0.2.0/mimvo/checks/security/injection/sentinel.py +44 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/__init__.py +50 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/_corpus.py +28 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/check.py +48 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/consistency.py +101 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/heuristic.py +453 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/hubness.py +109 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/nli.py +127 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/perplexity.py +140 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/probe.py +115 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/raguard.py +149 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/revprag.py +51 -0
- mimvo-0.2.0/mimvo/checks/security/poisoning/trustrag.py +192 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/__init__.py +86 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/check.py +51 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/detect_secrets.py +118 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/entropy.py +156 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/gitleaks.py +79 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/gliner2_pii.py +184 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/gliner_pii.py +150 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/heuristic.py +151 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/piiranha.py +77 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/presidio.py +161 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/starpii.py +54 -0
- mimvo-0.2.0/mimvo/checks/security/secrets/verify.py +62 -0
- mimvo-0.2.0/mimvo/cli.py +556 -0
- mimvo-0.2.0/mimvo/client.py +205 -0
- mimvo-0.2.0/mimvo/config.py +45 -0
- mimvo-0.2.0/mimvo/context.py +107 -0
- mimvo-0.2.0/mimvo/corpus.py +708 -0
- mimvo-0.2.0/mimvo/engine.py +350 -0
- mimvo-0.2.0/mimvo/exceptions.py +56 -0
- mimvo-0.2.0/mimvo/integrations/__init__.py +16 -0
- mimvo-0.2.0/mimvo/integrations/hook.py +216 -0
- mimvo-0.2.0/mimvo/mask.py +87 -0
- mimvo-0.2.0/mimvo/models/__init__.py +36 -0
- mimvo-0.2.0/mimvo/models/enums.py +171 -0
- mimvo-0.2.0/mimvo/models/finding.py +76 -0
- mimvo-0.2.0/mimvo/models/memory.py +79 -0
- mimvo-0.2.0/mimvo/models/results.py +366 -0
- mimvo-0.2.0/mimvo/owasp.py +39 -0
- mimvo-0.2.0/mimvo/plugins.py +142 -0
- mimvo-0.2.0/mimvo/py.typed +0 -0
- mimvo-0.2.0/mimvo/reporters/__init__.py +17 -0
- mimvo-0.2.0/mimvo/reporters/html.py +900 -0
- mimvo-0.2.0/mimvo/reporters/markdown.py +267 -0
- mimvo-0.2.0/mimvo/reporters/sarif.py +222 -0
- mimvo-0.2.0/mimvo/reporters/terminal.py +177 -0
- mimvo-0.2.0/mimvo/rules.py +486 -0
- mimvo-0.2.0/mimvo/sources/__init__.py +33 -0
- mimvo-0.2.0/mimvo/sources/base.py +261 -0
- mimvo-0.2.0/mimvo/sources/chroma.py +141 -0
- mimvo-0.2.0/mimvo/sources/jsonl.py +85 -0
- mimvo-0.2.0/mimvo/sources/langchain.py +251 -0
- mimvo-0.2.0/mimvo/sources/langgraph.py +137 -0
- mimvo-0.2.0/mimvo/sources/mem0.py +238 -0
- mimvo-0.2.0/mimvo/sources/pgvector.py +160 -0
- mimvo-0.2.0/mimvo/sources/pinecone.py +154 -0
- mimvo-0.2.0/mimvo/sources/qdrant.py +140 -0
- mimvo-0.2.0/mimvo/telemetry/__init__.py +116 -0
- mimvo-0.2.0/mimvo/telemetry/otel.py +79 -0
- mimvo-0.2.0/mimvo/text.py +469 -0
- mimvo-0.2.0/mimvo/vectors.py +155 -0
- mimvo-0.2.0/pyproject.toml +175 -0
mimvo-0.2.0/.gitignore
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
1
|
+
# Byte-compiled / optimized
|
|
2
|
+
__pycache__/
|
|
3
|
+
*.py[cod]
|
|
4
|
+
*$py.class
|
|
5
|
+
*.so
|
|
6
|
+
|
|
7
|
+
# Distribution / packaging
|
|
8
|
+
.Python
|
|
9
|
+
build/
|
|
10
|
+
dist/
|
|
11
|
+
downloads/
|
|
12
|
+
eggs/
|
|
13
|
+
.eggs/
|
|
14
|
+
*.egg-info/
|
|
15
|
+
*.egg
|
|
16
|
+
wheels/
|
|
17
|
+
share/python-wheels/
|
|
18
|
+
|
|
19
|
+
# Virtual environments
|
|
20
|
+
.venv/
|
|
21
|
+
venv/
|
|
22
|
+
env/
|
|
23
|
+
ENV/
|
|
24
|
+
|
|
25
|
+
# Tooling caches
|
|
26
|
+
.mypy_cache/
|
|
27
|
+
.ruff_cache/
|
|
28
|
+
.pytest_cache/
|
|
29
|
+
.hypothesis/
|
|
30
|
+
.coverage
|
|
31
|
+
htmlcov/
|
|
32
|
+
coverage.xml
|
|
33
|
+
|
|
34
|
+
# Editors / OS
|
|
35
|
+
.idea/
|
|
36
|
+
.vscode/
|
|
37
|
+
*.swp
|
|
38
|
+
.DS_Store
|
|
39
|
+
|
|
40
|
+
# Mimvo local artifacts (example and CLI outputs)
|
|
41
|
+
report.html
|
|
42
|
+
results.sarif
|
|
43
|
+
summary.md
|
|
44
|
+
findings.json
|
|
45
|
+
*.jsonl
|
|
46
|
+
!examples/**/*.jsonl
|
|
47
|
+
.benchmark-cache/
|
|
48
|
+
|
|
49
|
+
# Local script that deletes and republishes the GitHub repo
|
|
50
|
+
repush-single-commit.sh
|
mimvo-0.2.0/CHANGELOG.md
ADDED
|
@@ -0,0 +1,258 @@
|
|
|
1
|
+
# Changelog
|
|
2
|
+
|
|
3
|
+
All notable changes to this project are documented here. The format is based on
|
|
4
|
+
[Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this project
|
|
5
|
+
adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
|
|
6
|
+
|
|
7
|
+
## [Unreleased]
|
|
8
|
+
|
|
9
|
+
## [0.2.0] - 2026-10-11
|
|
10
|
+
|
|
11
|
+
First release published to PyPI (`pip install mimvo`). 0.1.0 was never
|
|
12
|
+
published; the notes below describe the changes since that version.
|
|
13
|
+
|
|
14
|
+
### Release
|
|
15
|
+
|
|
16
|
+
- Releases are published to PyPI from a `v*` tag by
|
|
17
|
+
`.github/workflows/release.yml`, using PyPI Trusted Publishing.
|
|
18
|
+
- The `mem0` extra now installs PyYAML, which `mimvo scan mem0 --config
|
|
19
|
+
file.yaml` needs.
|
|
20
|
+
|
|
21
|
+
### Changed: one engine contract, one rule catalogue, one finding model
|
|
22
|
+
|
|
23
|
+
A design pass to match how mature scanners are built. The scan logic and
|
|
24
|
+
the findings it produces are unchanged (same codes, severities, actions,
|
|
25
|
+
and fingerprints), but several public names moved. To upgrade:
|
|
26
|
+
|
|
27
|
+
| Before | After |
|
|
28
|
+
|---|---|
|
|
29
|
+
| `from mimvo.scan import ChromaScanSource, ...` | `from mimvo.sources import ChromaScanSource, ...` |
|
|
30
|
+
| `from mimvo.scan import render_html, render_markdown, render_sarif` | `from mimvo.reporters import ...` |
|
|
31
|
+
| `mimvo.scan.source` (helpers such as `to_record`) | `mimvo.sources.base` |
|
|
32
|
+
| `mimvo.scan.mask` | `mimvo.mask` |
|
|
33
|
+
| `mimvo.models.results.format_scan_summary` | `mimvo.reporters.format_scan_summary` (`str(report)` still works) |
|
|
34
|
+
| `mimvo.scan.owasp` | `mimvo.owasp` |
|
|
35
|
+
| `ScanFinding.id` / `.type` / `.action` (and the JSON keys) | `.record_id` / `.code` / `.recommended_action` |
|
|
36
|
+
| JSON `schema_version` `2.0` | `3.0` (the three renames above) |
|
|
37
|
+
| `SecurityCheck.specs` of `FindingSpec` | `SecurityCheck.rules` of `mimvo.rules.Rule` |
|
|
38
|
+
| `WriteGuard.allow(content, **kwargs)` | `allow(content, *, metadata=, memory_id=, neighbours=)` |
|
|
39
|
+
| A `BaseDetector` overriding `detect` without `needs_corpus` ran in the corpus pass | Raises `ConfigurationError`; set `needs_corpus = True` or `False` |
|
|
40
|
+
| Missing optional extra raised `ConfigurationError` | Raises `IntegrationError` |
|
|
41
|
+
| A failing store or hosted-API call raised `ConfigurationError` | Raises `BackendError` |
|
|
42
|
+
|
|
43
|
+
Both `IntegrationError` and `BackendError` subclass `MimvoError`, so code
|
|
44
|
+
catching `MimvoError` is unaffected.
|
|
45
|
+
|
|
46
|
+
#### Added
|
|
47
|
+
|
|
48
|
+
- **Batch detection.** Records reach checks in batches of
|
|
49
|
+
`Config(batch_size=64)`. Hugging Face detectors classify every chunk of a
|
|
50
|
+
batch in one pipeline call; `LakeraGuardDetector` and
|
|
51
|
+
`PromptShieldDetector` keep `concurrency=8` requests in flight. Custom
|
|
52
|
+
detectors can override `detect_batch` or set `concurrency`. A failure on
|
|
53
|
+
one record stays on that record (`BatchError`).
|
|
54
|
+
- **Plugins.** Packages can register checks (`mimvo.checks`) and scan
|
|
55
|
+
sources (`mimvo.sources`) as entry points. `mimvo scan ... --check NAME`,
|
|
56
|
+
`mimvo scan plugin NAME --option key=value`, and `mimvo plugins`.
|
|
57
|
+
- **Custom rules render properly.** Defining a `SecurityCheck` registers its
|
|
58
|
+
rules, so its findings show their own title and remediation in HTML,
|
|
59
|
+
Markdown, and SARIF instead of a generic "custom" entry.
|
|
60
|
+
- **Two-pass custom checks.** Any `MemoryCheck` can split work across the
|
|
61
|
+
stream and corpus passes with `phases()`, `observe()`, and `conclude()`;
|
|
62
|
+
the engine no longer special-cases `SecurityCheck`.
|
|
63
|
+
- `WriteGuard.inspect(..., neighbours=[...])` compares a new memory with
|
|
64
|
+
nearby stored ones (clusters, contradictions) without letting their own
|
|
65
|
+
findings affect the decision.
|
|
66
|
+
- `ScanReport.raise_for_errors()` raises `CheckError` for an incomplete scan.
|
|
67
|
+
- `WriteDecision.findings` are the scan findings themselves, so they keep
|
|
68
|
+
the checks' evidence.
|
|
69
|
+
|
|
70
|
+
### Renamed: MemorySec is now Mimvo
|
|
71
|
+
|
|
72
|
+
The project, the PyPI distribution, the import package, and the CLI are now
|
|
73
|
+
called **Mimvo** ([mimvo.dev](https://mimvo.dev)). Nothing else about
|
|
74
|
+
scanning behaviour changed. To upgrade:
|
|
75
|
+
|
|
76
|
+
| Before | After |
|
|
77
|
+
|---|---|
|
|
78
|
+
| `pip install memorysec` | `pip install mimvo` |
|
|
79
|
+
| `memorysec scan ...` | `mimvo scan ...` |
|
|
80
|
+
| `import memorysec` / `from memorysec import ...` | `import mimvo` / `from mimvo import ...` |
|
|
81
|
+
| `MemorySec`, `AsyncMemorySec` | `Mimvo`, `AsyncMimvo` |
|
|
82
|
+
| `MemorySecError` | `MimvoError` |
|
|
83
|
+
| `ScanReport.memorysec_version` (JSON `memorysec_version`) | `ScanReport.mimvo_version` (JSON `mimvo_version`) |
|
|
84
|
+
| JSON `schema_version` `1.2` | `2.0` (only the field rename above) |
|
|
85
|
+
| SARIF `partialFingerprints.memorysecFingerprint`, tool name `MemorySec` | `mimvoFingerprint`, tool name `Mimvo` |
|
|
86
|
+
| OpenTelemetry span `memorysec.scan` | `mimvo.scan` |
|
|
87
|
+
|
|
88
|
+
Fingerprint values are unchanged (they hash the rule id and record id only),
|
|
89
|
+
so existing tickets and suppression lists still match. GitHub code scanning
|
|
90
|
+
will treat the renamed SARIF fingerprint key as new alerts once.
|
|
91
|
+
|
|
92
|
+
### Since 0.1.0
|
|
93
|
+
|
|
94
|
+
Version 0.1.0 was an SDK that sat between an agent and its memory backend
|
|
95
|
+
(the project was called MemTrust then). Since then it has become a scanner:
|
|
96
|
+
it reads a memory store, reports poisoned facts, hidden instructions, and
|
|
97
|
+
leaked secrets, and recommends what to do with each record. It never writes
|
|
98
|
+
to the store. Everything below is relative to 0.1.0.
|
|
99
|
+
|
|
100
|
+
### Added
|
|
101
|
+
|
|
102
|
+
- **Store scanning.** `Mimvo().scan(source)` and `mimvo scan
|
|
103
|
+
chroma|qdrant|pgvector|pinecone|jsonl|langchain|mem0`. Sources are
|
|
104
|
+
read-only and stream records in batches; `--sample N` caps a scan.
|
|
105
|
+
- **LangChain, LangGraph, and mem0 sources.** `LangChainScanSource` reads
|
|
106
|
+
LangChain vector stores (InMemoryVectorStore, Chroma, Qdrant, Pinecone,
|
|
107
|
+
FAISS, PGVector, or any store by id with `get_by_ids`);
|
|
108
|
+
`LangGraphStoreScanSource` reads a LangGraph `BaseStore` (LangGraph and
|
|
109
|
+
LangMem long-term memory); `Mem0ScanSource` reads mem0's open-source
|
|
110
|
+
`Memory` (whole store or one user/agent/run) and the hosted
|
|
111
|
+
`MemoryClient`. CLI: `scan langchain --factory module:attr`,
|
|
112
|
+
`scan mem0 --config FILE | --api-key KEY`. Extras: `langchain`,
|
|
113
|
+
`langgraph`, `mem0`.
|
|
114
|
+
- **Reports.** A self-contained HTML report (verdict, triage plan,
|
|
115
|
+
findings with masked excerpts, fix steps, OWASP and CWE links), JSON
|
|
116
|
+
(`--json`, schema `2.0`), SARIF 2.1.0 (`--sarif`) for GitHub code
|
|
117
|
+
scanning, and Markdown (`--markdown`) for job summaries and PR comments.
|
|
118
|
+
Every format uses the rule catalogue in `mimvo.rules` and stable
|
|
119
|
+
per-finding `fingerprint`s.
|
|
120
|
+
- **CI gates.** `--fail-on SEVERITY` exits `1` when a finding reaches that
|
|
121
|
+
severity; `--min-confidence SCORE` ignores weaker findings for that gate;
|
|
122
|
+
`-q` prints counts only.
|
|
123
|
+
- **Confidence.** Every finding has a `confidence` from 0 to 1: the
|
|
124
|
+
combined score of the detectors that agreed on it (`1 - Π(1 - score)`).
|
|
125
|
+
A finding below 0.6 is reported one severity step lower.
|
|
126
|
+
- **Detectors.** Each check runs a list of detectors and can require
|
|
127
|
+
several to agree (`min_detectors`). Opt-in detectors: Hugging Face
|
|
128
|
+
classifiers (Prompt Guard 2, Protect AI and deepset DeBERTa, Sentinel,
|
|
129
|
+
Piiranha, StarPII), GLiNER, Presidio, detect-secrets, Azure Prompt
|
|
130
|
+
Shields, Lakera Guard, entropy, known-answer and DataSentinel, embedding
|
|
131
|
+
classifiers, perplexity and RAGuard, TrustRAG, hubness, embedding
|
|
132
|
+
consistency, temporal NLI, probe queries, and live secret verification.
|
|
133
|
+
- **Scan errors.** `ScanReport.errors`, `complete`, and
|
|
134
|
+
`records_with_errors` say which checks or detectors failed and on how
|
|
135
|
+
many records.
|
|
136
|
+
- **`WriteGuard` and `RetrieveGuard`** (`mimvo.integrations`): run the
|
|
137
|
+
scan on a memory before it is written, or on records just retrieved for a
|
|
138
|
+
query. They replace 0.1.0's `check_write` / `check_read`.
|
|
139
|
+
- **Accuracy benchmark** (`benchmarks/`): the default detectors measured on
|
|
140
|
+
public datasets they were not written against (GitHub's help articles,
|
|
141
|
+
AgentDojo, BIPIA, InjecAgent, PoisonedRAG), with a held-out test split.
|
|
142
|
+
Results are in `benchmarks/RESULTS.md`.
|
|
143
|
+
- **`mimvo.corpus`**: the packed record store and shared
|
|
144
|
+
nearest-neighbour table corpus detectors use, with a pluggable
|
|
145
|
+
`NeighbourIndex`. The `fast` extra installs numpy for the exact index.
|
|
146
|
+
- OpenTelemetry tracing (`mimvo[otel]`) and loguru logging; Mimvo
|
|
147
|
+
never adds or removes log sinks.
|
|
148
|
+
|
|
149
|
+
### Changed
|
|
150
|
+
|
|
151
|
+
- **Default detectors are scored and read context.** The injection and
|
|
152
|
+
poisoning heuristics judge each match by its sentence: help-text framing
|
|
153
|
+
("To disable 2FA, go to Settings"), conditionals, questions, limited
|
|
154
|
+
scope ("guest Wi-Fi", "when on VPN"), advice, and quotations lower the
|
|
155
|
+
score; sensitive assets, change markers, and directives aimed at the agent
|
|
156
|
+
raise it; a prohibition ("never disable MFA") drops the match. Bare
|
|
157
|
+
"review", "login", and "password" count only weakly. New coverage:
|
|
158
|
+
paraphrased overrides, typos in trigger words, base64/hex/URL/HTML/
|
|
159
|
+
reversed payloads, fake role tokens, task hijacking, authority spoofing,
|
|
160
|
+
limit removal, data exfiltration, secrets pushed into channels, and links
|
|
161
|
+
or falsehoods pushed into the agent's replies.
|
|
162
|
+
- **Severities follow one rule.** Explicit attack text is high
|
|
163
|
+
(`persistent_instruction`, `memory_poisoning`, `destination_redirect`);
|
|
164
|
+
model or behavioural evidence is medium (`known_answer`,
|
|
165
|
+
`embedding_injection`, `adversarial_text`, `embedding_mismatch`,
|
|
166
|
+
`temporal_contradiction`, `retrieval_flip`, `pii_detected`); corpus
|
|
167
|
+
statistics are low (`poisoning_cluster`, `hub_record`); a leaked secret
|
|
168
|
+
is critical.
|
|
169
|
+
- **A failing detector marks the scan incomplete.** It used to become a
|
|
170
|
+
critical `check_error` finding with action `delete` on every record. Now
|
|
171
|
+
the healthy detectors' results are kept, the failure is listed in
|
|
172
|
+
`report.errors`, no finding is created, and the CLI exits `2` (unless
|
|
173
|
+
`--allow-incomplete`). SARIF sets `executionSuccessful: false`; the HTML
|
|
174
|
+
and Markdown reports say the scan is incomplete. With `fail_closed=True`
|
|
175
|
+
(the default) `WriteGuard` refuses and `RetrieveGuard` drops records that
|
|
176
|
+
were not fully checked.
|
|
177
|
+
- **Scans stream.** Records are no longer collected into one list. Text
|
|
178
|
+
detectors run as records arrive; corpus detectors (TrustRAG, hubness,
|
|
179
|
+
NLI) run afterwards against packed float32 vectors and one shared k-NN
|
|
180
|
+
table. 50,000 records with 1,536-dimension vectors scan in about 90
|
|
181
|
+
seconds and 1.1 GB. TrustRAG's `embed=` path embeds each text once per
|
|
182
|
+
scan, and its lexical fallback compares records that share rare word
|
|
183
|
+
pairs instead of every pair.
|
|
184
|
+
- `AsyncMimvo.scan` runs in a worker thread so it no longer blocks the
|
|
185
|
+
event loop, and accepts async iterables.
|
|
186
|
+
- `Mimvo(checks=[...])` replaces a default check with the same name.
|
|
187
|
+
|
|
188
|
+
### Removed
|
|
189
|
+
|
|
190
|
+
- The write/read SDK: `check_write`, `check_read`, `protect`,
|
|
191
|
+
`ProtectedMemory`, `AsyncProtectedMemory`, `revoke_source`, `Decision`,
|
|
192
|
+
`ReadResult`, `SafeMemory`, `FilteredMemory`, `RevocationReport`, gate
|
|
193
|
+
actions (`allow`, `allow_with_warning`, `block`), modes (`observe`,
|
|
194
|
+
`warn`, `enforce`), and the `mimvo check` and `mimvo audit`
|
|
195
|
+
commands. Use `Mimvo.scan`, or `WriteGuard` / `RetrieveGuard` on the
|
|
196
|
+
write and read paths.
|
|
197
|
+
- Backends and adapters: `MemoryBackend`, `InMemoryBackend`,
|
|
198
|
+
`FunctionBackend`, the Mem0 and LangGraph adapters, and the `zep`
|
|
199
|
+
extra. Scan sources are the only store integrations.
|
|
200
|
+
- Correctness checks (contradiction, duplication, freshness,
|
|
201
|
+
generalization), source/scope/tenant/authority models and checks,
|
|
202
|
+
policies, the audit trail, secret redaction, the injectable clock, the
|
|
203
|
+
provenance model, and the semantic analyzer.
|
|
204
|
+
- The `check_error` finding code.
|
|
205
|
+
|
|
206
|
+
### Fixed
|
|
207
|
+
|
|
208
|
+
- `mimvo scan chroma` failed on any collection with stored vectors:
|
|
209
|
+
chromadb returns embeddings as a numpy array, which the source tested
|
|
210
|
+
with `or`.
|
|
211
|
+
- pgvector vectors returned as text (`"[0.1,0.2]"`, without the pgvector
|
|
212
|
+
adapter) were dropped, which silently turned the vector detectors off.
|
|
213
|
+
- `HubnessDetector` cached its table without its `k`, so two hubness
|
|
214
|
+
detectors with different `k` shared one result.
|
|
215
|
+
- `WriteGuard` put a detector name in `Finding.check`.
|
|
216
|
+
- A reset link after "password:" was reported as a leaked password.
|
|
217
|
+
|
|
218
|
+
## [0.1.0] - 2026-09-12
|
|
219
|
+
|
|
220
|
+
Not published to PyPI. Initial version. Mimvo v0.1 is an SDK/MVP: a vendor-neutral trust
|
|
221
|
+
boundary between an AI agent and its memory backend.
|
|
222
|
+
|
|
223
|
+
### Added
|
|
224
|
+
|
|
225
|
+
- **Facade**: `Mimvo` and `AsyncMimvo` with `check_write`, `check_read`,
|
|
226
|
+
`protect`, and `revoke_source`.
|
|
227
|
+
- **Domain models** (Pydantic v2): `Source`, `Scope`, `MemoryCandidate`,
|
|
228
|
+
`MemoryRecord`, `Finding`, `Decision`, `Provenance`, `Policy`, plus enums for
|
|
229
|
+
trust, severity, risk, action, status, and memory relationships.
|
|
230
|
+
- **Authority vs. confidence**: modeled as distinct fields, with authority
|
|
231
|
+
derived from source trust by default.
|
|
232
|
+
- **Security checks**: persistent-instruction/injection detection, secret
|
|
233
|
+
detection with redaction, tenant isolation, scope-promotion detection, and a
|
|
234
|
+
low-authority policy-write check (`untrusted_policy_write`).
|
|
235
|
+
- **Correctness checks**: freshness/expiry, local duplicate detection,
|
|
236
|
+
contradiction vs. supersession, and a generalization extension point.
|
|
237
|
+
- **Core enforcement** that cannot be bypassed by customizing the check list:
|
|
238
|
+
cross-tenant/cross-user read isolation, tenant-match on write, and filtering
|
|
239
|
+
of revoked/expired/quarantined/superseded records.
|
|
240
|
+
- **Policies**: declarative `Policy` (`when`/`require`) plus plain callables.
|
|
241
|
+
- **Semantic strategy**: deterministic `HeuristicSemanticAnalyzer` (default,
|
|
242
|
+
offline) and an optional provider-agnostic `LLMSemanticAnalyzer` with
|
|
243
|
+
timeouts and fail-open/closed behavior.
|
|
244
|
+
- **Backends**: `MemoryBackend`/`AsyncMemoryBackend` Protocols and in-memory
|
|
245
|
+
reference implementations; adapters for Mem0 and LangGraph `BaseStore`; a
|
|
246
|
+
generic `FunctionBackend`.
|
|
247
|
+
- **Audit**: JSON-serializable `AuditEvent`, `InMemoryAuditStore`, and
|
|
248
|
+
`JSONLAuditStore`. Secrets are redacted from audit content.
|
|
249
|
+
- **Telemetry**: stdlib `logging` (no `basicConfig`) and optional OpenTelemetry
|
|
250
|
+
tracing (injected, never global).
|
|
251
|
+
- **Modes**: `observe`, `warn`, `enforce` (default).
|
|
252
|
+
- **CLI**: `mimvo check` and `mimvo audit`.
|
|
253
|
+
- Test suite (unit, integration, security-invariant, property-based) and six
|
|
254
|
+
runnable examples.
|
|
255
|
+
|
|
256
|
+
[Unreleased]: https://github.com/jawadhussein462/mimvo/compare/v0.2.0...HEAD
|
|
257
|
+
[0.2.0]: https://github.com/jawadhussein462/mimvo/releases/tag/v0.2.0
|
|
258
|
+
[0.1.0]: https://github.com/jawadhussein462/mimvo/releases/tag/v0.1.0
|
mimvo-0.2.0/LICENSE
ADDED
|
@@ -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 Derivative
|
|
95
|
+
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, excluding
|
|
103
|
+
those notices that do not pertain to any part of the Derivative
|
|
104
|
+
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 do
|
|
117
|
+
not modify the License. You may add Your own attribution notices
|
|
118
|
+
within Derivative Works that You distribute, alongside or as an
|
|
119
|
+
addendum to the NOTICE text from the Work, provided that such
|
|
120
|
+
additional attribution notices cannot be construed as modifying
|
|
121
|
+
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 [yyyy] [name of copyright owner]
|
|
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.
|