deepintshield 2.3.0__tar.gz → 2.5.0__tar.gz

{deepintshield-2.3.0/src/deepintshield.egg-info → deepintshield-2.5.0}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: deepintshield
-Version: 2.3.0
-Summary: Unified Python SDK for routing chat, RAG, agentic tool-gating, identity, and MCP traffic through DeepintShield — drop-in across the top agentic frameworks.
+Version: 2.5.0
+Summary: Unified Python SDK for routing chat, RAG, agentic tool-gating, identity, and MCP traffic through DeepintShield - drop-in across the top agentic frameworks.
 Author: DeepintShield
 License: Apache-2.0
 Project-URL: Homepage, https://app.deepintshield.com
@@ -71,7 +71,7 @@ Dynamic: license-file
 
 # deepintshield
 
-Unified Python SDK for DeepintShield — one import, any provider, any agent framework.
+Unified Python SDK for DeepintShield - one import, any provider, any agent framework.
 
 `deepintshield` lets you keep writing idiomatic OpenAI / Anthropic / Bedrock /
 Google GenAI code **and** native agent-framework code (LangGraph, CrewAI,
@@ -79,20 +79,20 @@ OpenAI Agents SDK, LlamaIndex, AutoGen, PydanticAI) while automatically routing
 traffic through the DeepintShield gateway for guardrails, RAG filtering, agentic
 tool control, and agent identity.
 
-You pass **only two things — a virtual key and a base URL.** Everything else —
-the Entra / ZeroID / OIDC identity binding, tenant, scopes, and policy — is
+You pass **only two things - a virtual key and a base URL.** Everything else -
+the Entra / ZeroID / OIDC identity binding, tenant, scopes, and policy - is
 discovered from the gateway automatically. No identity GUIDs ever appear in your
 code.
 
 Protection comes in two layers:
 
-- **Transparent** — point any framework's *native* client at the gateway
+- **Transparent** - point any framework's *native* client at the gateway
   (`base_url` + a one-line header injector). Chat, embeddings, input/output and
   RAG-prompt guardrails, observability and identity all run server-side with
   **zero changes** to your agent code. Works for Python frameworks *and* any
   OpenAI-compatible platform (n8n, Flowise, Dify, raw HTTP).
-- **Enforcement** — one wrapping line adds local **tool gating** (DENY /
-  MASK / human approval) and **chunk-level RAG filtering** — the parts that
+- **Enforcement** - one wrapping line adds local **tool gating** (DENY /
+  MASK / human approval) and **chunk-level RAG filtering** - the parts that
   physically can't be done at the wire because the tool runs in your process.
 
 Traffic defaults to **`https://app.deepintshield.com`**. Override the gateway
@@ -126,7 +126,7 @@ pip install 'deepintshield[all]'                # everything
 
 ```bash
 export DEEPINTSHIELD_VIRTUAL_KEY="sk-..."
-# Optional — point at a self-hosted or staging gateway.
+# Optional - point at a self-hosted or staging gateway.
 export DEEPINTSHIELD_BASE_URL="https://gateway.example.com"
 ```
 
@@ -251,7 +251,7 @@ allowed, raw = shield.rag.filter(query="What's the badge rule?", chunks=chunks)
 ### Guard a framework retriever (post-retrieval, one line)
 
 Wrap any LangChain / LlamaIndex retriever so unauthorised chunks are dropped
-*after* retrieval and *before* they reach the LLM — ACL/provenance filtering
+*after* retrieval and *before* they reach the LLM - ACL/provenance filtering
 your retriever can't do itself:
 
 ```python
@@ -269,7 +269,7 @@ embedder.embed_query("text")                              # raises if the gatewa
 ```
 
 > If you obtain your embedder from a framework binder (below), input-side
-> screening already happens server-side — `guard_embedder` is for embedders you
+> screening already happens server-side - `guard_embedder` is for embedders you
 > don't route through the gateway.
 
 ---
@@ -292,7 +292,7 @@ shield.autogen().model_client("gpt-4o-mini")     # OpenAIChatCompletionClient
 shield.pydanticai().model("gpt-4o-mini")
 ```
 
-Framework-agnostic primitives — wire any SDK, in any language-compatible
+Framework-agnostic primitives - wire any SDK, in any language-compatible
 client, by hand:
 
 ```python
@@ -302,23 +302,23 @@ headers = shield.create_headers(provider="anthropic")   # Portkey-style header i
 ```
 
 > **Universal:** anything that speaks OpenAI-compatible HTTP gets transparent
-> protection with *zero code* — set the Base URL to `shield.endpoint("openai")`
+> protection with *zero code* - set the Base URL to `shield.endpoint("openai")`
 > and the key to your VK. That covers n8n, Flowise, Dify and raw HTTP, not just
 > Python.
 
 ## Agentic tool enforcement
 
 The part a base URL can't do: gate **local** tool execution. Each gated call
-runs `decide()` first and the verdict maps to a Python outcome — `ALLOW` runs
+runs `decide()` first and the verdict maps to a Python outcome - `ALLOW` runs
 the body, `MASK` redacts PII kwargs, `REQUIRE_APPROVAL` blocks for a human, and
 `DENY` raises `GuardrailDenied`.
 
-### Zero extra code — enforcement is automatic and non-bypassable
+### Zero extra code - enforcement is automatic and non-bypassable
 
 The moment you construct `DeepintShield(...)`, the SDK installs guards for every
 agent framework you've imported (LangGraph, CrewAI, LlamaIndex, AutoGen,
 PydanticAI, the OpenAI Agents SDK, LiteLLM). After that, **compiling / building
-an agent yields an already-governed object** — you can't forget to gate it and
+an agent yields an already-governed object** - you can't forget to gate it and
 you can't bypass it by invoking the un-governed one (there isn't one).
 
 ```python
@@ -334,7 +334,7 @@ g = StateGraph(State)
 g.add_node("read_step", crm_read)
 g.add_node("admin_step", admin_grant)
 ...
-app = g.compile()                      # auto-governed — every node now gated by the PDP
+app = g.compile()                      # auto-governed - every node now gated by the PDP
 
 app.invoke({...})                      # a DENY raises GuardrailDenied before the node runs
 ```
@@ -345,32 +345,32 @@ If you import a framework *after* building the client, call
 > **Security follows the implementation, not the label.** Each node/tool is
 > governed by its **function name** (`crm_read`), not the node label
 > (`read_step`), and the decision is bound to a fingerprint of the function's
-> **source** — so editing the body is detected and policies target `crm_read`.
+> **source** - so editing the body is detected and policies target `crm_read`.
 >
 > **Trust boundary:** these client guards are cooperative defense-in-depth. A
 > determined process can un-patch them or call a tool's raw function, so the
 > gateway (MCP / LLM in the call path) remains the authoritative boundary.
 
-### `govern()` — register + threat-scan + instrument (explicit, idempotent)
+### `govern()` - register + threat-scan + instrument (explicit, idempotent)
 
 `govern()` does everything the auto-guard does, explicitly: it **describes** the
 agent's declared tool surface, **registers** that blueprint with the server
-(which **threat-scans each tool's source** for RCE / shell-out / exfiltration —
-OWASP Agentic **T11 / T17** — server-side, ZDR), and **instruments** every call.
+(which **threat-scans each tool's source** for RCE / shell-out / exfiltration -
+OWASP Agentic **T11 / T17** - server-side, ZDR), and **instruments** every call.
 
 ```python
-app = shield.agentic.govern(app)       # idempotent — safe alongside the auto-guard
+app = shield.agentic.govern(app)       # idempotent - safe alongside the auto-guard
 ```
 
 A tool whose source scans malicious is flagged (Agentic → Findings) and, when
-the workspace enables **Enforce code threat** (Rollout), denied — even if a
+the workspace enables **Enforce code threat** (Rollout), denied - even if a
 policy would otherwise allow it. A tool called but never declared shows up as
 **ASI04 drift** under Agentic → Discovery.
 
-### `guard()` — LangChain callback / in-place instrument
+### `guard()` - LangChain callback / in-place instrument
 
 `shield.agentic.guard()` (no argument) returns a native LangChain callback
-handler; attaching it once gates *every* tool the agent calls — the framework
+handler; attaching it once gates *every* tool the agent calls - the framework
 supplies the tool name and the **gateway resolves the tier, policy, recovery
 cost and identity server-side**.
 
@@ -382,7 +382,7 @@ agent_executor.invoke({"input": "…"}, config={"callbacks": [guard]})
 `guard(target)` auto-detects and instruments a framework object in place:
 
 ```python
-shield.agentic.guard(compiled_graph)   # LangGraph — gate every tool node
+shield.agentic.guard(compiled_graph)   # LangGraph - gate every tool node
 shield.agentic.guard(crewai_tools)     # CrewAI BaseTools
 shield.agentic.guard(openai_agent)     # OpenAI Agents FunctionTools
 shield.agentic.guard(pydantic_agent)   # PydanticAI agent
@@ -435,7 +435,7 @@ When the virtual key is bound to an identity provider, the SDK auto-discovers
 the binding (`GET /api/agentic-security/vk-credential-info`), selects the right
 credential (Entra Agent ID FIC / ZeroID RFC 8693 / generic OIDC), and attaches a
 fresh `X-Agent-Token` on every decision. On Azure compute the Managed Identity
-is detected automatically — **no GUIDs, authority, or scopes in your code.**
+is detected automatically - **no GUIDs, authority, or scopes in your code.**
 
 ```python
 info = shield.agentic.credential_info     # ops visibility into the binding
@@ -473,7 +473,7 @@ app = graph.compile()
 ## Multimodal guardrails (transparent)
 
 Image generation, image edits, audio (TTS / transcription), video, embedding and
-rerank requests are guarded **at the gateway** — no SDK changes and no extra code.
+rerank requests are guarded **at the gateway** - no SDK changes and no extra code.
 Keep using the native provider SDKs through DeepIntShield; when the operator
 enables `GUARDRAILS_MULTIMODAL`, the gateway evaluates the text these requests
 already carry (image/TTS/video prompts, transcripts) and the binary artifacts
@@ -482,7 +482,7 @@ themselves, blocking or flagging per your policies.
 ```python
 client = shield.openai()
 
-# Guarded automatically — the image prompt is evaluated before generation.
+# Guarded automatically - the image prompt is evaluated before generation.
 img = client.images.generate(model="gpt-image-1", prompt="a serene mountain lake")
 
 # A blocked prompt surfaces as the provider SDK's normal HTTP error:
@@ -501,7 +501,7 @@ enforcing (`sync`) or observe-only (`shadow`).
 
 ## MCP
 
-Generic MCP support — works with any server connected to your DeepintShield
+Generic MCP support - works with any server connected to your DeepintShield
 gateway. No per-server SDK code; the same `Tool` / `MCPClient` API serves
 DeepWiki, Context7, GitHub MCP, an internal one, and so on.
 
@@ -596,7 +596,7 @@ print(AgentExecutor(agent=agent, tools=mcp_tools).invoke(
 )["output"])
 ```
 
-LangGraph reuses the same `mcp_tools` list — drop them into a `ToolNode`.
+LangGraph reuses the same `mcp_tools` list - drop them into a `ToolNode`.
 
 ---
 
@@ -605,16 +605,16 @@ LangGraph reuses the same `mcp_tools` list — drop them into a `ToolNode`.
 The SDK automatically participates in the gateway's two cost-reduction layers
 (both controlled by workspace switches under **Cost Optimization**):
 
-- **Provider prompt caching** — every chat client returned by `shield.openai()`,
+- **Provider prompt caching** - every chat client returned by `shield.openai()`,
   `shield.anthropic()`, etc. ships an `httpx` request hook that injects
   Anthropic `cache_control` markers and an OpenAI `prompt_cache_key` so the
   provider reuses KV state for the static prompt prefix. Cached tokens come
   back at the provider's reduced rate (50% off OpenAI, 90% off Anthropic).
-- **Gemini context caching** — opt in with `shield.genai_cached()` (drop-in
+- **Gemini context caching** - opt in with `shield.genai_cached()` (drop-in
   for `shield.genai()`). The wrapper manages the `cachedContents` resource
   lifecycle behind the scenes; the first call with a new static prefix runs
   normally and the next call within the TTL window reuses the cache.
-- **Semantic caching** — runs on the gateway; short-circuits requests whose
+- **Semantic caching** - runs on the gateway; short-circuits requests whose
   embeddings match a previous response within the configured similarity
   threshold. The SDK doesn't need any code change to benefit; results flow
   back through the normal API.

{deepintshield-2.3.0 → deepintshield-2.5.0}/README.md

@@ -1,6 +1,6 @@
 # deepintshield
 
-Unified Python SDK for DeepintShield — one import, any provider, any agent framework.
+Unified Python SDK for DeepintShield - one import, any provider, any agent framework.
 
 `deepintshield` lets you keep writing idiomatic OpenAI / Anthropic / Bedrock /
 Google GenAI code **and** native agent-framework code (LangGraph, CrewAI,
@@ -8,20 +8,20 @@ OpenAI Agents SDK, LlamaIndex, AutoGen, PydanticAI) while automatically routing
 traffic through the DeepintShield gateway for guardrails, RAG filtering, agentic
 tool control, and agent identity.
 
-You pass **only two things — a virtual key and a base URL.** Everything else —
-the Entra / ZeroID / OIDC identity binding, tenant, scopes, and policy — is
+You pass **only two things - a virtual key and a base URL.** Everything else -
+the Entra / ZeroID / OIDC identity binding, tenant, scopes, and policy - is
 discovered from the gateway automatically. No identity GUIDs ever appear in your
 code.
 
 Protection comes in two layers:
 
-- **Transparent** — point any framework's *native* client at the gateway
+- **Transparent** - point any framework's *native* client at the gateway
   (`base_url` + a one-line header injector). Chat, embeddings, input/output and
   RAG-prompt guardrails, observability and identity all run server-side with
   **zero changes** to your agent code. Works for Python frameworks *and* any
   OpenAI-compatible platform (n8n, Flowise, Dify, raw HTTP).
-- **Enforcement** — one wrapping line adds local **tool gating** (DENY /
-  MASK / human approval) and **chunk-level RAG filtering** — the parts that
+- **Enforcement** - one wrapping line adds local **tool gating** (DENY /
+  MASK / human approval) and **chunk-level RAG filtering** - the parts that
   physically can't be done at the wire because the tool runs in your process.
 
 Traffic defaults to **`https://app.deepintshield.com`**. Override the gateway
@@ -55,7 +55,7 @@ pip install 'deepintshield[all]'                # everything
 
 ```bash
 export DEEPINTSHIELD_VIRTUAL_KEY="sk-..."
-# Optional — point at a self-hosted or staging gateway.
+# Optional - point at a self-hosted or staging gateway.
 export DEEPINTSHIELD_BASE_URL="https://gateway.example.com"
 ```
 
@@ -180,7 +180,7 @@ allowed, raw = shield.rag.filter(query="What's the badge rule?", chunks=chunks)
 ### Guard a framework retriever (post-retrieval, one line)
 
 Wrap any LangChain / LlamaIndex retriever so unauthorised chunks are dropped
-*after* retrieval and *before* they reach the LLM — ACL/provenance filtering
+*after* retrieval and *before* they reach the LLM - ACL/provenance filtering
 your retriever can't do itself:
 
 ```python
@@ -198,7 +198,7 @@ embedder.embed_query("text")                              # raises if the gatewa
 ```
 
 > If you obtain your embedder from a framework binder (below), input-side
-> screening already happens server-side — `guard_embedder` is for embedders you
+> screening already happens server-side - `guard_embedder` is for embedders you
 > don't route through the gateway.
 
 ---
@@ -221,7 +221,7 @@ shield.autogen().model_client("gpt-4o-mini")     # OpenAIChatCompletionClient
 shield.pydanticai().model("gpt-4o-mini")
 ```
 
-Framework-agnostic primitives — wire any SDK, in any language-compatible
+Framework-agnostic primitives - wire any SDK, in any language-compatible
 client, by hand:
 
 ```python
@@ -231,23 +231,23 @@ headers = shield.create_headers(provider="anthropic")   # Portkey-style header i
 ```
 
 > **Universal:** anything that speaks OpenAI-compatible HTTP gets transparent
-> protection with *zero code* — set the Base URL to `shield.endpoint("openai")`
+> protection with *zero code* - set the Base URL to `shield.endpoint("openai")`
 > and the key to your VK. That covers n8n, Flowise, Dify and raw HTTP, not just
 > Python.
 
 ## Agentic tool enforcement
 
 The part a base URL can't do: gate **local** tool execution. Each gated call
-runs `decide()` first and the verdict maps to a Python outcome — `ALLOW` runs
+runs `decide()` first and the verdict maps to a Python outcome - `ALLOW` runs
 the body, `MASK` redacts PII kwargs, `REQUIRE_APPROVAL` blocks for a human, and
 `DENY` raises `GuardrailDenied`.
 
-### Zero extra code — enforcement is automatic and non-bypassable
+### Zero extra code - enforcement is automatic and non-bypassable
 
 The moment you construct `DeepintShield(...)`, the SDK installs guards for every
 agent framework you've imported (LangGraph, CrewAI, LlamaIndex, AutoGen,
 PydanticAI, the OpenAI Agents SDK, LiteLLM). After that, **compiling / building
-an agent yields an already-governed object** — you can't forget to gate it and
+an agent yields an already-governed object** - you can't forget to gate it and
 you can't bypass it by invoking the un-governed one (there isn't one).
 
 ```python
@@ -263,7 +263,7 @@ g = StateGraph(State)
 g.add_node("read_step", crm_read)
 g.add_node("admin_step", admin_grant)
 ...
-app = g.compile()                      # auto-governed — every node now gated by the PDP
+app = g.compile()                      # auto-governed - every node now gated by the PDP
 
 app.invoke({...})                      # a DENY raises GuardrailDenied before the node runs
 ```
@@ -274,32 +274,32 @@ If you import a framework *after* building the client, call
 > **Security follows the implementation, not the label.** Each node/tool is
 > governed by its **function name** (`crm_read`), not the node label
 > (`read_step`), and the decision is bound to a fingerprint of the function's
-> **source** — so editing the body is detected and policies target `crm_read`.
+> **source** - so editing the body is detected and policies target `crm_read`.
 >
 > **Trust boundary:** these client guards are cooperative defense-in-depth. A
 > determined process can un-patch them or call a tool's raw function, so the
 > gateway (MCP / LLM in the call path) remains the authoritative boundary.
 
-### `govern()` — register + threat-scan + instrument (explicit, idempotent)
+### `govern()` - register + threat-scan + instrument (explicit, idempotent)
 
 `govern()` does everything the auto-guard does, explicitly: it **describes** the
 agent's declared tool surface, **registers** that blueprint with the server
-(which **threat-scans each tool's source** for RCE / shell-out / exfiltration —
-OWASP Agentic **T11 / T17** — server-side, ZDR), and **instruments** every call.
+(which **threat-scans each tool's source** for RCE / shell-out / exfiltration -
+OWASP Agentic **T11 / T17** - server-side, ZDR), and **instruments** every call.
 
 ```python
-app = shield.agentic.govern(app)       # idempotent — safe alongside the auto-guard
+app = shield.agentic.govern(app)       # idempotent - safe alongside the auto-guard
 ```
 
 A tool whose source scans malicious is flagged (Agentic → Findings) and, when
-the workspace enables **Enforce code threat** (Rollout), denied — even if a
+the workspace enables **Enforce code threat** (Rollout), denied - even if a
 policy would otherwise allow it. A tool called but never declared shows up as
 **ASI04 drift** under Agentic → Discovery.
 
-### `guard()` — LangChain callback / in-place instrument
+### `guard()` - LangChain callback / in-place instrument
 
 `shield.agentic.guard()` (no argument) returns a native LangChain callback
-handler; attaching it once gates *every* tool the agent calls — the framework
+handler; attaching it once gates *every* tool the agent calls - the framework
 supplies the tool name and the **gateway resolves the tier, policy, recovery
 cost and identity server-side**.
 
@@ -311,7 +311,7 @@ agent_executor.invoke({"input": "…"}, config={"callbacks": [guard]})
 `guard(target)` auto-detects and instruments a framework object in place:
 
 ```python
-shield.agentic.guard(compiled_graph)   # LangGraph — gate every tool node
+shield.agentic.guard(compiled_graph)   # LangGraph - gate every tool node
 shield.agentic.guard(crewai_tools)     # CrewAI BaseTools
 shield.agentic.guard(openai_agent)     # OpenAI Agents FunctionTools
 shield.agentic.guard(pydantic_agent)   # PydanticAI agent
@@ -364,7 +364,7 @@ When the virtual key is bound to an identity provider, the SDK auto-discovers
 the binding (`GET /api/agentic-security/vk-credential-info`), selects the right
 credential (Entra Agent ID FIC / ZeroID RFC 8693 / generic OIDC), and attaches a
 fresh `X-Agent-Token` on every decision. On Azure compute the Managed Identity
-is detected automatically — **no GUIDs, authority, or scopes in your code.**
+is detected automatically - **no GUIDs, authority, or scopes in your code.**
 
 ```python
 info = shield.agentic.credential_info     # ops visibility into the binding
@@ -402,7 +402,7 @@ app = graph.compile()
 ## Multimodal guardrails (transparent)
 
 Image generation, image edits, audio (TTS / transcription), video, embedding and
-rerank requests are guarded **at the gateway** — no SDK changes and no extra code.
+rerank requests are guarded **at the gateway** - no SDK changes and no extra code.
 Keep using the native provider SDKs through DeepIntShield; when the operator
 enables `GUARDRAILS_MULTIMODAL`, the gateway evaluates the text these requests
 already carry (image/TTS/video prompts, transcripts) and the binary artifacts
@@ -411,7 +411,7 @@ themselves, blocking or flagging per your policies.
 ```python
 client = shield.openai()
 
-# Guarded automatically — the image prompt is evaluated before generation.
+# Guarded automatically - the image prompt is evaluated before generation.
 img = client.images.generate(model="gpt-image-1", prompt="a serene mountain lake")
 
 # A blocked prompt surfaces as the provider SDK's normal HTTP error:
@@ -430,7 +430,7 @@ enforcing (`sync`) or observe-only (`shadow`).
 
 ## MCP
 
-Generic MCP support — works with any server connected to your DeepintShield
+Generic MCP support - works with any server connected to your DeepintShield
 gateway. No per-server SDK code; the same `Tool` / `MCPClient` API serves
 DeepWiki, Context7, GitHub MCP, an internal one, and so on.
 
@@ -525,7 +525,7 @@ print(AgentExecutor(agent=agent, tools=mcp_tools).invoke(
 )["output"])
 ```
 
-LangGraph reuses the same `mcp_tools` list — drop them into a `ToolNode`.
+LangGraph reuses the same `mcp_tools` list - drop them into a `ToolNode`.
 
 ---
 
@@ -534,16 +534,16 @@ LangGraph reuses the same `mcp_tools` list — drop them into a `ToolNode`.
 The SDK automatically participates in the gateway's two cost-reduction layers
 (both controlled by workspace switches under **Cost Optimization**):
 
-- **Provider prompt caching** — every chat client returned by `shield.openai()`,
+- **Provider prompt caching** - every chat client returned by `shield.openai()`,
   `shield.anthropic()`, etc. ships an `httpx` request hook that injects
   Anthropic `cache_control` markers and an OpenAI `prompt_cache_key` so the
   provider reuses KV state for the static prompt prefix. Cached tokens come
   back at the provider's reduced rate (50% off OpenAI, 90% off Anthropic).
-- **Gemini context caching** — opt in with `shield.genai_cached()` (drop-in
+- **Gemini context caching** - opt in with `shield.genai_cached()` (drop-in
   for `shield.genai()`). The wrapper manages the `cachedContents` resource
   lifecycle behind the scenes; the first call with a new static prefix runs
   normally and the next call within the TTL window reuses the cache.
-- **Semantic caching** — runs on the gateway; short-circuits requests whose
+- **Semantic caching** - runs on the gateway; short-circuits requests whose
   embeddings match a previous response within the configured similarity
   threshold. The SDK doesn't need any code change to benefit; results flow
   back through the normal API.

{deepintshield-2.3.0 → deepintshield-2.5.0}/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "deepintshield"
-version = "2.3.0"
-description = "Unified Python SDK for routing chat, RAG, agentic tool-gating, identity, and MCP traffic through DeepintShield — drop-in across the top agentic frameworks."
+version = "2.5.0"
+description = "Unified Python SDK for routing chat, RAG, agentic tool-gating, identity, and MCP traffic through DeepintShield - drop-in across the top agentic frameworks."
 readme = "README.md"
 requires-python = ">=3.10"
 license = { text = "Apache-2.0" }

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/__init__.py

@@ -1,5 +1,5 @@
 """
-DeepintShield — unified Python SDK.
+DeepintShield - unified Python SDK.
 
 Quick start
 -----------

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/_gemini_cache.py

@@ -1,6 +1,6 @@
 """Gemini context-cache management for the DeepintShield SDK.
 
-Google Gemini doesn't have inline cache markers like Anthropic — instead it
+Google Gemini doesn't have inline cache markers like Anthropic - instead it
 exposes an explicit ``cachedContents`` resource with its own lifecycle:
 
 1. Client calls ``client.caches.create(...)`` with the static portion of the
@@ -14,23 +14,23 @@ exposes an explicit ``cachedContents`` resource with its own lifecycle:
 This module wraps that lifecycle so customers don't have to manage cache
 resources by hand. Design choices for zero-latency / scalable / parallel use:
 
-* **Per-process registry** — a thread-safe dict of ``prefix_hash → (name,
+* **Per-process registry** - a thread-safe dict of ``prefix_hash → (name,
   expires_at_ms)``. Lookups are O(1), no I/O.
-* **Fire-and-forget creation** — the first request that sees a prefix doesn't
+* **Fire-and-forget creation** - the first request that sees a prefix doesn't
   wait for the cache resource to exist. It runs through the provider as a
   normal call, while a background thread creates the cache for the *next*
   call. This keeps the first call's latency identical to no-cache.
-* **Lazy TTL eviction** — expired entries are dropped at lookup time. No
+* **Lazy TTL eviction** - expired entries are dropped at lookup time. No
   background sweeper, no extra timers.
-* **Workspace switch respected** — the gateway strips ``cached_content`` from
+* **Workspace switch respected** - the gateway strips ``cached_content`` from
   outbound requests when the workspace has prompt caching disabled or
   ``google`` not in the provider allow-list, so the SDK's injection becomes
   a no-op end-to-end.
-* **Caller-aware** — if the caller already passed ``cached_content`` on a
+* **Caller-aware** - if the caller already passed ``cached_content`` on a
   call's config, we leave it alone.
 
 Phase 4 ships an explicit ``shield.gemini_cache_manager()`` plus a wrapped
-``shield.genai_cached()`` client that calls into it automatically — opt-in
+``shield.genai_cached()`` client that calls into it automatically - opt-in
 because Gemini's storage cost (~$1/M tok/hr on 1.5 Pro) makes blanket
 auto-caching a footgun for prefixes that don't repeat often.
 """
@@ -64,7 +64,7 @@ def _ttl_seconds_from_string(ttl: str | int | None) -> int:
     """Parse a workspace-style TTL string ("5m", "1h", "6h", "24h") into seconds.
 
     Falls through to ``DEFAULT_TTL_SECONDS`` on unknown values so misconfigured
-    workspaces never throw at request time — they just use the default.
+    workspaces never throw at request time - they just use the default.
     """
     if ttl is None:
         return DEFAULT_TTL_SECONDS

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/_prompt_cache.py

@@ -55,7 +55,7 @@ def _mark_anthropic_block_cacheable(block: Any, marker: dict[str, str]) -> bool:
     if not isinstance(block, dict):
         return False
     if "cache_control" in block:
-        # Caller already marked this block — respect it.
+        # Caller already marked this block - respect it.
         return False
     block["cache_control"] = marker
     return True
@@ -76,7 +76,7 @@ def _inject_anthropic(
     if "system" in requested:
         system = body.get("system")
         if isinstance(system, list) and system:
-            # System is already a list of blocks — mark the last one.
+            # System is already a list of blocks - mark the last one.
             _mark_anthropic_block_cacheable(system[-1], marker)
         elif isinstance(system, str) and system.strip():
             # Promote string-form system prompts into block form so we can attach
@@ -115,7 +115,7 @@ def _compute_prefix_hash(
 ) -> str:
     """Stable 16-char SHA-256 prefix of (system messages + tool defs).
 
-    OpenAI uses this as a cache partition key — same hash → same cache bucket.
+    OpenAI uses this as a cache partition key - same hash → same cache bucket.
     Identical prefixes from different conversations therefore share cache
     entries. Returns "" when there's no meaningful static prefix to hash.
     """
@@ -200,7 +200,7 @@ def build_request_hook(
         encoded = json.dumps(body, separators=(",", ":")).encode("utf-8")
         request.headers["content-length"] = str(len(encoded))
         # httpx routes the actual transmitted bytes through `request.stream`
-        # — a single-shot byte iterator built when the Request was first
+        # - a single-shot byte iterator built when the Request was first
         # constructed. Updating only `_content` leaves the stream pointing
         # at the original bytes, so on the wire httpx emits the *old* body
         # while advertising the *new* content-length. The result on the
@@ -211,7 +211,7 @@ def build_request_hook(
         request._content = encoded  # type: ignore[attr-defined]
         try:
             from httpx._content import ByteStream  # type: ignore
-        except ImportError:  # httpx <0.24 fallback — module path drifted.
+        except ImportError:  # httpx <0.24 fallback - module path drifted.
             from httpx import _content  # type: ignore
             ByteStream = getattr(_content, "ByteStream", None)
         if ByteStream is not None:

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/__init__.py

@@ -3,7 +3,7 @@
 The tool-enforcement half of the SDK: ``decide()`` runs before a gated tool
 executes and the verdict (ALLOW / DENY / MASK / REQUIRE_APPROVAL) maps to a
 return value or an exception. All Entra / identity / policy machinery is
-auto-discovered from the gateway — the user passes only ``virtual_key`` and
+auto-discovered from the gateway - the user passes only ``virtual_key`` and
 ``base_url`` to :class:`~deepintshield.client.DeepintShield`.
 
 Typically consumed via ``shield.agentic`` rather than imported directly::

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/credentials/__init__.py

@@ -3,7 +3,7 @@
 A credential is anything that can produce an agent identity token to present
 alongside the platform VK on every PEP request. The right implementation is
 auto-selected by :class:`AgenticEngine` based on the ``provider_type`` it
-discovers from the gateway — developers almost never instantiate these
+discovers from the gateway - developers almost never instantiate these
 directly.
 """
 

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/credentials/base.py

@@ -1,4 +1,4 @@
-"""AgentCredential — the protocol every identity provider implementation
+"""AgentCredential - the protocol every identity provider implementation
 follows. The engine only ever calls ``get_token()`` on the abstract type;
 the concrete provider (Entra / ZeroID / OIDC) handles the wire details.
 """
@@ -32,7 +32,7 @@ class StaticAgentCredential:
     """Returns a pre-baked token verbatim. For local dev / tests where you
     want a known-good token without running an FIC exchange.
 
-    Do not use in production — the token won't refresh.
+    Do not use in production - the token won't refresh.
     """
 
     def __init__(self, token: str, provider_type: str = "static") -> None:

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/credentials/entra.py

@@ -1,11 +1,11 @@
-"""Entra Agent ID credential — Federated Identity Credential exchange.
+"""Entra Agent ID credential - Federated Identity Credential exchange.
 
 Flow per call (only when the cached token is missing or near expiry):
 
     1. Get a self-token from the local Managed Identity for audience
        ``api://AzureADTokenExchange``. Uses DefaultAzureCredential so the MI
        binding is auto-detected from the host (AKS workload identity, App
-       Service identity, VM identity, etc.) — no client_id needs to appear
+       Service identity, VM identity, etc.) - no client_id needs to appear
        in user code.
 
     2. POST that as the ``client_assertion`` to the blueprint's token
@@ -43,7 +43,7 @@ class EntraAgentCredential:
     """Concrete AgentCredential for Microsoft Entra Agent ID blueprints.
 
     Constructed via :class:`AgenticEngine` from the values discovered at
-    GET /api/agentic-security/vk-credential-info — developers should not
+    GET /api/agentic-security/vk-credential-info - developers should not
     instantiate this directly except for overrides.
     """
 
@@ -87,7 +87,7 @@ class EntraAgentCredential:
         if cached:
             return cached
         with self._cache.lock():
-            # Re-check inside lock — another thread may have refreshed.
+            # Re-check inside lock - another thread may have refreshed.
             cached = self._cache.get()
             if cached:
                 return cached

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/credentials/oidc.py

@@ -1,4 +1,4 @@
-"""Generic OIDC credential — Client-Credentials grant against any OIDC IdP
+"""Generic OIDC credential - Client-Credentials grant against any OIDC IdP
 that publishes a JWKS endpoint (Okta, Auth0, Keycloak, …).
 
 Used when the platform VK is bound to a ``generic_oidc`` identity provider on

{deepintshield-2.3.0 → deepintshield-2.5.0}/src/deepintshield/agentic/credentials/zeroid.py

@@ -1,9 +1,9 @@
-"""ZeroID credential — RFC 8693 token exchange.
+"""ZeroID credential - RFC 8693 token exchange.
 
 ZeroID's wire format follows OAuth 2.0 Token Exchange (RFC 8693) with a
 standard ``act`` claim chain and SPIFFE/WIMSE identity URIs. The gateway's
 broker verifies the resulting JWT against ZeroID's JWKS the same way it does
-for Entra — the SDK side is just the exchange dance.
+for Entra - the SDK side is just the exchange dance.
 
 For a full integration you'll typically swap in the vendor SDK's own token
 provider here; this implementation is the minimal HTTP shape that works for