blocklog 0.2.0__tar.gz → 0.2.4__tar.gz

{blocklog-0.2.0 → blocklog-0.2.4}/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.4
 Name: blocklog
-Version: 0.2.0
+Version: 0.2.4
 Summary: Infrastructure for AI Decision-Making — record, replay, verify, and govern AI agent decisions
-Author-email: Blocklog <contact@blockloghq.com>
-Maintainer-email: Blocklog <contact@blockloghq.com>
+Author-email: Blocklog <founder@blocklogsecurity.com>
+Maintainer-email: Blocklog <founder@blocklogsecurity.com>
 License: MIT
 Project-URL: Homepage, https://blockloghq.com
 Project-URL: Documentation, https://docs.blockloghq.com
@@ -25,12 +25,31 @@ Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: httpx<1.0,>=0.27
 Requires-Dist: pydantic<3.0,>=2.8
-Requires-Dist: requests>=2.32.0
+Provides-Extra: requests
+Requires-Dist: requests>=2.32.0; extra == "requests"
+Provides-Extra: langchain
+Requires-Dist: langchain-core>=0.2.0; extra == "langchain"
+Provides-Extra: langgraph
+Requires-Dist: langchain-core>=0.2.0; extra == "langgraph"
+Requires-Dist: langgraph>=0.2.0; extra == "langgraph"
+Provides-Extra: openai
+Requires-Dist: openai>=1.0.0; extra == "openai"
+Provides-Extra: litellm
+Requires-Dist: litellm>=1.40.0; extra == "litellm"
+Provides-Extra: all
+Requires-Dist: langchain-core>=0.2.0; extra == "all"
+Requires-Dist: langgraph>=0.2.0; extra == "all"
+Requires-Dist: openai>=1.0.0; extra == "all"
+Requires-Dist: litellm>=1.40.0; extra == "all"
 Provides-Extra: dev
 Requires-Dist: pytest>=8.0.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.23.0; extra == "dev"
 Requires-Dist: build>=1.0.0; extra == "dev"
 Requires-Dist: twine>=5.0.0; extra == "dev"
+Requires-Dist: langchain-core>=0.2.0; extra == "dev"
+Requires-Dist: langgraph>=0.2.0; extra == "dev"
+Requires-Dist: openai>=1.0.0; extra == "dev"
+Requires-Dist: litellm>=1.40.0; extra == "dev"
 Dynamic: license-file
 
 # Blocklog Python SDK
@@ -56,7 +75,7 @@ When AI agents execute actions in production, the context behind their decisions
 - **Trace tool calls**, capturing inputs and outputs seamlessly.
 - **Record decisions** with detailed metadata, inputs, and outputs.
 - **Request approvals** via non-blocking human-in-the-loop workflows.
-- **Verify records** cryptographically against a blockchain anchor.
+- **Verify records** cryptographically against Ed25519 signatures.
 - **Generate compliance evidence** (e.g., SOC2, GDPR).
 - **Investigate failures with replay** using forensic timelines and root-cause analysis.
 
@@ -128,6 +147,53 @@ if __name__ == "__main__":
     run_agent()
 ```
 
+## Signup And Teams
+
+```python
+from blocklog import (
+    AsyncBlocklogClient,
+    BlocklogClient,
+    can_manage_members,
+    can_manage_team,
+    get_primary_team,
+    is_team_owner,
+)
+
+client = BlocklogClient(
+    base_url="https://your-blocklog-host/api/v1",
+)
+
+signup = client.auth.signup(
+    username="jane",
+    email="jane@example.com",
+    password="ChangeMe123!",
+    workspace_name="Acme Security",
+)
+
+print(signup.team.name)
+print(is_team_owner(signup.team, signup.user.user_id))
+print(signup.team.owner_user_id)
+
+client.set_access_token(signup.token)
+
+teams = client.teams.list()
+primary_team = get_primary_team(teams)
+
+if primary_team and can_manage_team(primary_team, signup.user.user_id):
+    client.teams.update(primary_team.id, default_sla_minutes=30)
+    members = client.teams.members.list(primary_team.id)
+    if members and can_manage_members(members[0]):
+        client.teams.notify_test(primary_team.id)
+
+
+async def async_example() -> None:
+    async_client = AsyncBlocklogClient(base_url="https://your-blocklog-host/api/v1")
+    login = await async_client.auth.login("jane@example.com", "ChangeMe123!")
+    async_client.set_access_token(login.token)
+    teams = await async_client.teams.list()
+    print(teams)
+```
+
 ---
 
 ## What Happens Under The Hood?
@@ -162,6 +228,7 @@ Blocklog seamlessly auto-instruments popular AI frameworks:
 - **LangChain**: Trace chains, LLMs, and tools automatically via `client.instrument_langchain()`.
 - **LangGraph**: Hook into graph states via `client.instrument_langgraph()`.
 - **OpenAI Agents**: Track official SDK executions via `client.instrument_openai_agents()`.
+- **LiteLLM**: Instrument any LiteLLM completion call via `client.instrument_litellm()`.
 
 Read more in the [Integrations Guide](docs/integrations.md).
 
@@ -173,7 +240,7 @@ Read more in the [Integrations Guide](docs/integrations.md).
 |-------|-------|
 | **Getting Started** | [Installation](docs/installation.md) • [Quickstart](docs/quickstart.md) • [Core Concepts](docs/concepts.md) |
 | **Instrumentation** | [Tracing](docs/tracing.md) • [Decisions](docs/decisions.md) • [Decorators](docs/decorators.md) |
-| **Frameworks** | [Integrations](docs/integrations.md) |
+| **Frameworks** | [Integrations](docs/integrations.md) — LangChain, LangGraph, OpenAI Agents, LiteLLM |
 | **Operations** | [Async Usage](docs/async.md) • [Production](docs/production.md) • [Performance](docs/performance.md) |
 | **Reference** | [API Reference](docs/api-reference.md) • [Error Handling](docs/error-handling.md) • [Troubleshooting](docs/troubleshooting.md) |
 | **Misc** | [Examples](docs/examples.md) • [Changelog](docs/changelog.md) |
@@ -197,6 +264,8 @@ We provide several runnable scripts in the `examples/` directory to help you und
 - **Retry Handling**: Built-in exponential backoff for transient network issues.
 - **Signing**: Optional Ed25519 payload signing for tamper-evident, cryptographically verifiable logs.
 - **Middleware Hooks**: Add custom logic to redact PII or inject metadata before payloads leave your server.
+- **Typed Team APIs**: Ownership-aware models for teams, members, and signup responses.
+- **Standardized Exceptions**: Authentication, authorization, validation, conflict, rate-limit, and server error mapping.
 
 Read our [Production Best Practices](docs/production.md) for more details.
 
@@ -215,6 +284,7 @@ Blocklog can be configured via environment variables:
 | Variable | Description | Default |
 |----------|-------------|---------|
 | `BLOCKLOG_API_KEY` | Your Blocklog API key | `""` |
+| `BLOCKLOG_ACCESS_TOKEN` | User access token for dashboard-style APIs | `""` |
 | `BLOCKLOG_BASE_URL` | API base URL | `http://127.0.0.1:8000/api/v1` |
 | `BLOCKLOG_SDK_SIGNING_KEY` | Optional seed key for hash signing | `""` |
 | `BLOCKLOG_TIMEOUT` | Request timeout in seconds | `10` |
@@ -257,6 +327,20 @@ If the SDK cannot connect to the Blocklog API:
 3. Ensure network connectivity to the API server
 4. Check firewall settings
 
+## Teams API
+
+```python
+from blocklog import BlocklogClient
+
+client = BlocklogClient(access_token="user-token")
+
+teams = client.teams.list()
+team = client.teams.get(teams[0].id)
+members = client.teams.members.list(team.id)
+result = client.teams.notify_test(team.id)
+print(result.results)
+```
+
 ## Links
 
 - **Homepage**: https://blockloghq.com

blocklog-0.2.0/src/blocklog.egg-info/PKG-INFO → blocklog-0.2.4/README.md

@@ -1,38 +1,3 @@
-Metadata-Version: 2.4
-Name: blocklog
-Version: 0.2.0
-Summary: Infrastructure for AI Decision-Making — record, replay, verify, and govern AI agent decisions
-Author-email: Blocklog <contact@blockloghq.com>
-Maintainer-email: Blocklog <contact@blockloghq.com>
-License: MIT
-Project-URL: Homepage, https://blockloghq.com
-Project-URL: Documentation, https://docs.blockloghq.com
-Project-URL: Repository, https://github.com/blockloghq/blocklog-python
-Project-URL: Issues, https://github.com/blockloghq/blocklog-python/issues
-Keywords: ai,agents,observability,governance,security,audit,replay
-Classifier: Development Status :: 3 - Alpha
-Classifier: Intended Audience :: Developers
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Classifier: Programming Language :: Python :: 3.13
-Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Classifier: Topic :: System :: Monitoring
-Requires-Python: >=3.10
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: httpx<1.0,>=0.27
-Requires-Dist: pydantic<3.0,>=2.8
-Requires-Dist: requests>=2.32.0
-Provides-Extra: dev
-Requires-Dist: pytest>=8.0.0; extra == "dev"
-Requires-Dist: pytest-asyncio>=0.23.0; extra == "dev"
-Requires-Dist: build>=1.0.0; extra == "dev"
-Requires-Dist: twine>=5.0.0; extra == "dev"
-Dynamic: license-file
-
 # Blocklog Python SDK
 
 **Record, audit, and investigate every decision your AI agents make.**
@@ -56,7 +21,7 @@ When AI agents execute actions in production, the context behind their decisions
 - **Trace tool calls**, capturing inputs and outputs seamlessly.
 - **Record decisions** with detailed metadata, inputs, and outputs.
 - **Request approvals** via non-blocking human-in-the-loop workflows.
-- **Verify records** cryptographically against a blockchain anchor.
+- **Verify records** cryptographically against Ed25519 signatures.
 - **Generate compliance evidence** (e.g., SOC2, GDPR).
 - **Investigate failures with replay** using forensic timelines and root-cause analysis.
 
@@ -128,6 +93,53 @@ if __name__ == "__main__":
     run_agent()
 ```
 
+## Signup And Teams
+
+```python
+from blocklog import (
+    AsyncBlocklogClient,
+    BlocklogClient,
+    can_manage_members,
+    can_manage_team,
+    get_primary_team,
+    is_team_owner,
+)
+
+client = BlocklogClient(
+    base_url="https://your-blocklog-host/api/v1",
+)
+
+signup = client.auth.signup(
+    username="jane",
+    email="jane@example.com",
+    password="ChangeMe123!",
+    workspace_name="Acme Security",
+)
+
+print(signup.team.name)
+print(is_team_owner(signup.team, signup.user.user_id))
+print(signup.team.owner_user_id)
+
+client.set_access_token(signup.token)
+
+teams = client.teams.list()
+primary_team = get_primary_team(teams)
+
+if primary_team and can_manage_team(primary_team, signup.user.user_id):
+    client.teams.update(primary_team.id, default_sla_minutes=30)
+    members = client.teams.members.list(primary_team.id)
+    if members and can_manage_members(members[0]):
+        client.teams.notify_test(primary_team.id)
+
+
+async def async_example() -> None:
+    async_client = AsyncBlocklogClient(base_url="https://your-blocklog-host/api/v1")
+    login = await async_client.auth.login("jane@example.com", "ChangeMe123!")
+    async_client.set_access_token(login.token)
+    teams = await async_client.teams.list()
+    print(teams)
+```
+
 ---
 
 ## What Happens Under The Hood?
@@ -162,6 +174,7 @@ Blocklog seamlessly auto-instruments popular AI frameworks:
 - **LangChain**: Trace chains, LLMs, and tools automatically via `client.instrument_langchain()`.
 - **LangGraph**: Hook into graph states via `client.instrument_langgraph()`.
 - **OpenAI Agents**: Track official SDK executions via `client.instrument_openai_agents()`.
+- **LiteLLM**: Instrument any LiteLLM completion call via `client.instrument_litellm()`.
 
 Read more in the [Integrations Guide](docs/integrations.md).
 
@@ -173,7 +186,7 @@ Read more in the [Integrations Guide](docs/integrations.md).
 |-------|-------|
 | **Getting Started** | [Installation](docs/installation.md) • [Quickstart](docs/quickstart.md) • [Core Concepts](docs/concepts.md) |
 | **Instrumentation** | [Tracing](docs/tracing.md) • [Decisions](docs/decisions.md) • [Decorators](docs/decorators.md) |
-| **Frameworks** | [Integrations](docs/integrations.md) |
+| **Frameworks** | [Integrations](docs/integrations.md) — LangChain, LangGraph, OpenAI Agents, LiteLLM |
 | **Operations** | [Async Usage](docs/async.md) • [Production](docs/production.md) • [Performance](docs/performance.md) |
 | **Reference** | [API Reference](docs/api-reference.md) • [Error Handling](docs/error-handling.md) • [Troubleshooting](docs/troubleshooting.md) |
 | **Misc** | [Examples](docs/examples.md) • [Changelog](docs/changelog.md) |
@@ -197,6 +210,8 @@ We provide several runnable scripts in the `examples/` directory to help you und
 - **Retry Handling**: Built-in exponential backoff for transient network issues.
 - **Signing**: Optional Ed25519 payload signing for tamper-evident, cryptographically verifiable logs.
 - **Middleware Hooks**: Add custom logic to redact PII or inject metadata before payloads leave your server.
+- **Typed Team APIs**: Ownership-aware models for teams, members, and signup responses.
+- **Standardized Exceptions**: Authentication, authorization, validation, conflict, rate-limit, and server error mapping.
 
 Read our [Production Best Practices](docs/production.md) for more details.
 
@@ -215,6 +230,7 @@ Blocklog can be configured via environment variables:
 | Variable | Description | Default |
 |----------|-------------|---------|
 | `BLOCKLOG_API_KEY` | Your Blocklog API key | `""` |
+| `BLOCKLOG_ACCESS_TOKEN` | User access token for dashboard-style APIs | `""` |
 | `BLOCKLOG_BASE_URL` | API base URL | `http://127.0.0.1:8000/api/v1` |
 | `BLOCKLOG_SDK_SIGNING_KEY` | Optional seed key for hash signing | `""` |
 | `BLOCKLOG_TIMEOUT` | Request timeout in seconds | `10` |
@@ -257,6 +273,20 @@ If the SDK cannot connect to the Blocklog API:
 3. Ensure network connectivity to the API server
 4. Check firewall settings
 
+## Teams API
+
+```python
+from blocklog import BlocklogClient
+
+client = BlocklogClient(access_token="user-token")
+
+teams = client.teams.list()
+team = client.teams.get(teams[0].id)
+members = client.teams.members.list(team.id)
+result = client.teams.notify_test(team.id)
+print(result.results)
+```
+
 ## Links
 
 - **Homepage**: https://blockloghq.com
@@ -269,4 +299,4 @@ If the SDK cannot connect to the Blocklog API:
 
 ## License
 
-MIT License. See the LICENSE file for details.
+MIT License. See the LICENSE file for details.

{blocklog-0.2.0 → blocklog-0.2.4}/docs/api-reference.md

@@ -1,6 +1,19 @@
+```markdown
 # API Reference
 
-This document provides a comprehensive API reference derived from the Blocklog SDK source code. It covers all public modules, method signatures, parameters, return types, exceptions, and usage examples.
+## Environment Variables
+
+| Variable | Required | Default | Description |
+|---|---|---|---|
+| `BLOCKLOG_API_KEY` | Yes | — | Your Blocklog API key |
+| `BLOCKLOG_BASE_URL` | No | `https://blocklogsecurity.com/api/v1` | API endpoint |
+| `BLOCKLOG_SDK_SIGNING_KEY` | No | `""` | HMAC-SHA256 signing key |
+| `BLOCKLOG_TIMEOUT` | No | `10` | Request timeout in seconds |
+| `BLOCKLOG_MAX_RETRIES` | No | `3` | Retry attempts on failure |
+| `BLOCKLOG_BATCH_SIZE` | No | `100` | Events per batch flush |
+| `BLOCKLOG_FLUSH_INTERVAL` | No | `2` | Seconds between auto-flushes |
+
+---
 
 ## Initialization
 
@@ -17,15 +30,17 @@ def init(
 ) -> "BlocklogClient"
 ```
 **Purpose:** Initializes the Blocklog SDK. Call once at application startup.
+
 **Parameters:**
 - `api_key` (str, optional): Your Blocklog API key. Falls back to `BLOCKLOG_API_KEY`.
 - `base_url` (str, optional): Override the default API base URL. Falls back to `BLOCKLOG_BASE_URL`.
-- `signing_key` (str, optional): Optional seed key for deterministic hash signing of log payloads. Falls back to `BLOCKLOG_SDK_SIGNING_KEY`. Note: This is NOT cryptographic Ed25519 signing - it generates a deterministic SHA256 hash for tamper-evidence purposes.
+- `signing_key` (str, optional): HMAC-SHA256 key for tamper-evident log signatures. Falls back to `BLOCKLOG_SDK_SIGNING_KEY`. Note: this is not asymmetric Ed25519 signing.
 - `timeout` (float, optional): Per-request timeout in seconds (default: 10). Falls back to `BLOCKLOG_TIMEOUT`.
 - `max_retries` (int, optional): Number of automatic retries (default: 3). Falls back to `BLOCKLOG_MAX_RETRIES`.
 - `debug` (bool, optional): If `True`, logs every outbound request to stderr.
+
 **Returns:** `BlocklogClient`
-**Exceptions:** None directly, though misconfigurations may cause failures later.
+
 **Example:**
 ```python
 client = blocklog.init(api_key="blk_live_...")
@@ -46,13 +61,18 @@ def agent(
     metadata: dict[str, Any] | None = None,
 )
 ```
-**Purpose:** Decorator that traces an AI agent function or class, linking it to a Blocklog session. Automatically handles async functions.
+**Purpose:** Traces an AI agent function, linking it to a Blocklog session. Automatically handles sync and async functions.
+
+> **Warning:** Class decoration only emits `AGENT_START`. For full lifecycle tracing, decorate the specific method (e.g. `run`, `execute`) rather than the class itself.
+
 **Parameters:**
-- `name`: Human-readable agent name.
+- `name`: Human-readable agent name. Defaults to `func.__name__`.
 - `version`: Semver-style version string.
 - `tags`: Optional list of string tags.
 - `metadata`: Arbitrary extra data.
+
 **Returns:** Decorated callable or class.
+
 **Example:**
 ```python
 @blocklog.agent(name="analyst", version="1.0")
@@ -60,6 +80,8 @@ def run_analyst():
     pass
 ```
 
+---
+
 ### `@blocklog.tool`
 ```python
 def tool(
@@ -71,17 +93,20 @@ def tool(
     metadata: dict[str, Any] | None = None,
 )
 ```
-**Purpose:** Decorator that records a tool call as a Blocklog event, tracking inputs, outputs, and duration. Inherits the trace from `@agent`. Handles async functions automatically.
+**Purpose:** Records a tool call as a Blocklog event, capturing inputs, outputs, duration, and errors. Inherits trace context from the surrounding `@agent`.
+
 **Parameters:**
-- `name`: Human-readable tool name.
+- `name`: Human-readable tool name. Defaults to `func.__name__`.
 - `schema`: Optional dict describing the input schema.
 - `tags`: Optional string tags.
 - `metadata`: Arbitrary extra data.
+
 **Returns:** Decorated callable.
+
 **Example:**
 ```python
 @blocklog.tool(name="fetch-price")
-def get_price(ticker: str):
+def get_price(ticker: str) -> float:
     return 100.0
 ```
 
@@ -101,56 +126,63 @@ def decision(
     trace_id: str | None = None,
 ) -> Generator[DecisionContext, None, None]
 ```
-**Purpose:** Context manager for recording an AI decision. Yields a `DecisionContext`.
+**Purpose:** Context manager for recording an AI decision.
+
 **Parameters:**
-- `type` (str): Decision type identifier (e.g. "BUY").
+- `type` (str): Decision type identifier (e.g. `"BUY"`).
 - `asset` (str, optional): Asset this decision is about.
 - `confidence` (float, optional): Model confidence score (0 to 1).
 - `metadata` (dict, optional): Extra fields.
+
 **Returns:** Yields `DecisionContext`.
+
 **Example:**
 ```python
 with blocklog.decision(type="BUY", asset="TSLA") as d:
+    d.record_input(price=412.50)
     d.record_output(status="executed")
 ```
 
 ### `DecisionContext`
 Live handle to a decision being recorded.
+
 - **`record_input(**kwargs)`**: Record structured inputs. Returns `self`.
 - **`record_output(**kwargs)`**: Record structured outputs. Returns `self`.
 - **`tag(*tags)`**: Attach string labels. Returns `self`.
-- **`request_approval(reason: str, reviewer: str | None = None)`**: Non-blocking request for human approval. Returns `self`.
-- **`verify() -> dict`**: Immediately verify the decision. Raises `RuntimeError` if called before the decision is committed (i.e. before the `with` block begins).
+- **`request_approval(reason, reviewer=None)`**: Non-blocking request for human approval. Returns `self`.
+- **`verify() -> dict`**: Verify the decision after the `with` block has exited. Raises `RuntimeError` if called before the decision is committed.
 
 ---
 
 ## Clients
 
 ### `BlocklogClient`
-**Purpose:** The synchronous core client. Usually instantiated via `blocklog.init()`.
+**Purpose:** Synchronous core client. Usually instantiated via `blocklog.init()`.
 
 **Methods:**
-- `from_env() -> BlocklogClient`: Class method to create a client solely from env vars.
-- `add_hook(hook: Callable) -> BlocklogClient`: Register a middleware hook.
-- `instrument_openai_agents() -> BlocklogClient`
-- `instrument_langchain() -> BlocklogClient`
-- `instrument_langgraph() -> BlocklogClient`
-- `event(event_type: str, payload: dict, **kwargs) -> IngestResponse`: Emit synchronous event.
-- `enqueue(event_type: str, payload: dict, **kwargs)`: Enqueue batched event.
-- `flush(batch=None)`: Flush event buffer.
+- `from_env() -> BlocklogClient`: Create a client from environment variables.
+- `add_hook(hook: Callable) -> BlocklogClient`: Register a middleware hook applied to every outbound event.
+- `instrument_openai_agents() -> BlocklogClient`: Auto-instrument the OpenAI Agents SDK.
+- `instrument_langchain() -> BlocklogClient`: Auto-instrument LangChain.
+- `instrument_langgraph() -> BlocklogClient`: Auto-instrument LangGraph.
+- `event(event_type, payload, **kwargs) -> IngestResponse`: Emit a single event immediately.
+- `enqueue(event_type, payload, **kwargs)`: Enqueue an event for batched delivery.
+- `flush(batch=None)`: Flush the event buffer.
 
 ### `AsyncBlocklogClient`
-**Purpose:** The asynchronous client for manual async flushing and event dispatch.
-**Methods:** Inherits from `BlocklogClient` but overrides network calls to be async:
-- `async event(event_type: str, payload: dict, **kwargs) -> IngestResponse`
+**Purpose:** Async client for use in async applications.
+
+**Methods:** Same as `BlocklogClient` with async overrides:
+- `async event(event_type, payload, **kwargs) -> IngestResponse`
 - `async flush(batch=None)`
 
 ### `BlocklogConfig`
 **Purpose:** Configuration Pydantic model.
+
 **Fields:**
 - `base_url`: str
 - `api_key`: str
-- `signing_key`: str (seed key for hash-based signing, NOT Ed25519)
+- `signing_key`: str — HMAC-SHA256 key, not Ed25519
 - `timeout`: float
 - `max_retries`: int
 - `batch_size`: int
@@ -160,47 +192,47 @@ Live handle to a decision being recorded.
 
 ## Layer 2 Sub-modules
 
-*(Available via `blocklog.module.*` or `client.module.*`)*
+*Available via `client.module.method()` or imported directly from `blocklog.api.*`.*
 
-### `blocklog.api.decisions.DecisionsClient`
-**Purpose:** Manage AI Decision records natively.
+### `blocklog.api.decisions`
 - `create(decision_type, asset, confidence, metadata, trace_id, session_id, agent_id) -> dict`
 - `list() -> list[dict]`
-- `get(decision_id: str) -> dict`
-- `verify(decision_id: str) -> dict`
-- `timeline(decision_id: str) -> list[dict]`
-- `evidence(decision_id: str) -> dict`
-- `replay(decision_id: str) -> dict`
+- `get(decision_id) -> dict`
+- `verify(decision_id) -> dict`
+- `timeline(decision_id) -> list[dict]`
+- `evidence(decision_id) -> dict`
+- `replay(decision_id) -> dict`
 
 ### `blocklog.approval`
-**Purpose:** Human-in-the-loop (HITL) workflows.
+Human-in-the-loop workflows.
 - `request(decision_id, reason, reviewer, log_id, metadata) -> dict`
 - `reject(reviewer, reason, decision_id) -> dict`
 - `escalate(from_reviewer, to_reviewer, reason) -> dict`
 - `list_overrides() -> list[dict]`
 - `audit_trail() -> list[dict]`
 
-### `blocklog.incident`
-**Purpose:** Incident management API.
+### `blocklog.api.incidents`
+Incident management. Not exported from the top-level `blocklog` namespace — import via `from blocklog.api.incidents import IncidentsClient` or access via `client.incidents`.
 - `create(title, trace_id, severity, description, metadata) -> IncidentHandle`
 - `get(incident_id) -> IncidentHandle`
 - `list_all() -> list[IncidentHandle]`
 
 ### `blocklog.replay`
-**Purpose:** Forensic replays.
+Forensic replay sessions.
 - `replay(trace_id, token_id, metadata) -> ReplaySession`
 
-### `blocklog.verify`
-**Purpose:** Cryptographic verification against blockchain anchors.
-- `log(log_id: str) -> dict`
-- `batch(batch_id: str) -> dict`
-- `decision(decision_id: str) -> dict`
+### `blocklog.api.verify`
+Access via `client.verify` to avoid naming conflict with `DecisionContext.verify()`.
+- `log(log_id) -> dict`
+- `batch(batch_id) -> dict`
+- `decision(decision_id) -> dict`
 
 ### `blocklog.compliance`
-**Purpose:** Compliance report generation.
+Compliance report generation.
 - `generate(trace_id, framework, date_from, date_to, metadata) -> dict`
 - `get(report_id) -> dict`
 - `list() -> list[dict]`
 - `dashboard() -> dict`
 - `share(report_id, expires_in, recipient_email) -> dict`
 - `export(report_id, download) -> dict`
+```

{blocklog-0.2.0 → blocklog-0.2.4}/docs/examples.md

@@ -26,7 +26,6 @@ Done. Check your Blocklog dashboard.
 A detailed execution trace printed to the terminal of the `stock-trader` agent checking the market price and placing an order. A decision record is produced and verified:
 ```
   ✓ Decision recorded: <uuid>
-  Verifying decision against blockchain anchor...
   ✓ Verification status: verified
 ```
 

{blocklog-0.2.0 → blocklog-0.2.4}/docs/index.md

@@ -2,7 +2,7 @@
 
 **Infrastructure for AI Decision-Making**
 
-Blocklog provides the infrastructure to record, audit, and investigate AI decisions. It allows you to wrap your AI agents, capture their context and tool calls, and record structured decisions that are securely anchored and verifiable.
+Blocklog provides the infrastructure to record, audit, and investigate AI decisions. It allows you to wrap your AI agents, capture their context and tool calls, and record structured decisions that are cryptographically signed and verifiable.
 
 ## What problem it solves
 
@@ -11,7 +11,7 @@ When AI agents make decisions in production—like executing a trade, granting a
 Blocklog solves this by:
 1. **Tracing AI Agents**: Capturing the exact state, tool inputs/outputs, and prompts.
 2. **Recording Decisions**: Creating a structured, auditable record of AI actions.
-3. **Providing Verifiability**: Anchoring decisions cryptographically.
+3. **Providing Verifiability**: Signing decisions cryptographically with Ed25519.
 4. **Enabling Human-in-the-Loop (HITL)**: Requesting approvals for high-stakes decisions seamlessly.
 
 ## Documentation Index

blocklog-0.2.4/docs/installation.md

@@ -0,0 +1,26 @@
+# Installation
+
+## Requirements
+
+Python 3.10 or newer.
+
+## Install
+
+```bash
+pip install blocklog
+```
+
+## Verify
+
+```python
+import blocklog
+print(blocklog.__version__)  # 0.2.2
+```
+
+## Development
+
+```bash
+git clone https://github.com/Blockloghq/blocklog-python.git
+cd blocklog-python
+pip install -e ".[dev]"
+```