Public API Index¶
Every name in the __all__ of promptise and its public subpackages in 1.6.0, with
a one-line summary and the page that explains it. Import a name from the module
it is listed under. Names that are also exported from the top-level promptise
package are listed there only. The API Reference pages
render the full docstrings.
promptise¶
Top level. from promptise import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
Action |
enum | Action to take when a finding is detected. | Security Guardrails |
AdaptiveLesson |
class | A stored lesson, as returned by list_lessons. |
Adaptive Strategy |
AdaptiveStrategyConfig |
class | Configuration for the adaptive strategy system. | Adaptive Strategy |
AdaptiveStrategyManager |
class | Manages adaptive strategy learning for an agent. | Adaptive Strategy |
AgentContext |
class | Unified context layer for an agent process. | Agent Context |
AgentEvent |
class | A structured notification event from the Promptise framework. | Agent Events & Notifications |
AgentIdentity |
class | The identity of an agent: for tracing and attributing its actions. | Agent Identity Architecture |
AgentProcess |
class | Lifecycle container for a long-running agent process. | Agent Processes |
AgentRuntime |
class | Multi-process agent runtime manager. | Runtime Manager |
AgentSection |
class | Agent-level configuration section. | SuperAgent Files |
approval_elicitation_callback |
function | Build an MCP elicitation_callback that asks an approval handler. |
Human-in-the-Loop Approval |
ApprovalDecision |
class | The decision on an approval request. | Human-in-the-Loop Approval |
ApprovalHandler |
class | Protocol for approval handlers. | Human-in-the-Loop Approval |
ApprovalPolicy |
class | Configuration for human-in-the-loop approval. | Human-in-the-Loop Approval |
ApprovalRequest |
class | A request for human approval of a tool call. | Human-in-the-Loop Approval |
ApprovalRule |
class | A single deny, ask or allow rule for the classifier. | AutoApprovalClassifier |
AutoApprovalClassifier |
class | Decision-hierarchy wrapper for approval handlers. | AutoApprovalClassifier |
BaseContext |
class | Extensible context container. | Prompt Context System |
BaseNode |
class | Concrete base class for graph nodes. | Reasoning Graph Nodes |
BudgetConfig |
class | Autonomy budget configuration (unknown keys warn and are ignored). | Runtime Configuration |
build_agent |
function | Build an MCP-first agent and return a PromptiseAgent. |
Building Agents |
build_superagent |
function | Build the agent in a .superagent file together with its whole team. |
SuperAgent Files |
CacheStats |
class | Cache performance statistics. | Semantic Cache |
CallbackApprovalHandler |
class | Approval handler that delegates to an async Python callable. | Human-in-the-Loop Approval |
CallbackSink |
class | Deliver events to a Python callable. | Agent Events & Notifications |
CallerContext |
class | Identity and metadata for the caller of an agent invocation. | Building Agents |
check_model |
function | Diagnose a model string without calling anything. | Models & Providers |
ChromaProvider |
class | Adapter for ChromaDB <https://www.trychroma.com/>_. |
Agent Memory |
Chunk |
class | A chunk of a document ready for embedding and retrieval. | RAG (Retrieval-Augmented Generation) |
Chunker |
class | Splits a document into chunks suitable for embedding. | RAG (Retrieval-Augmented Generation) |
ClassifierDecisionTrace |
class | Diagnostic record of which layer decided a request. | Human-in-the-Loop Approval |
ClassifierStats |
class | Cumulative counts of decisions by hierarchy layer. | AutoApprovalClassifier |
classify_failure |
function | Classify an error as infrastructure, strategy, or unknown. | Adaptive Strategy |
content_hash |
function | Return a short deterministic hash of a text string. | RAG (Retrieval-Augmented Generation) |
ContentSafetyDetector |
class | Detect harmful content across 13 safety categories. | Security Guardrails |
ContextEngine |
class | Unified context assembly with token budgeting. | Context Engine |
ConversationMessage |
class | A single conversation message. | Conversation Persistence |
ConversationStore |
class | Interface for conversation persistence backends. | Conversation Persistence |
CredentialCategory |
enum | Credential detection categories. Pass a set of these to PromptiseSecurityScanner(enable_credentials={...}). |
Security Guardrails |
CredentialDetector |
class | Detect leaked credentials using 96 regex patterns from gitleaks/trufflehog. | Security Guardrails |
CrossAgent |
class | Metadata wrapper for a peer agent to be exposed as a tool. | Cross-Agent Delegation |
CrossAgentConfig |
class | Cross-agent reference configuration. | SuperAgent Files |
CustomRule |
class | A developer-defined regex detection rule. | Security Guardrails |
DEFAULT_DESTRUCTIVE_VERBS |
constant | tuple of 63 values | AutoApprovalClassifier |
DEFAULT_READ_ONLY_PREFIXES |
constant | tuple of 15 values | AutoApprovalClassifier |
DelegationError |
exception | A cross-agent call was refused before it reached the peer. | Cross-Agent Delegation |
DetailedModelConfig |
class | Detailed model configuration with provider-specific parameters. | SuperAgent Files |
Document |
class | A raw document loaded from a data source. | RAG (Retrieval-Augmented Generation) |
DocumentLoader |
class | Fetches raw documents from a data source. | RAG (Retrieval-Augmented Generation) |
DoneEvent |
class | The agent has finished processing. | Streaming with Tool Visibility |
Edge |
class | A directed edge between two nodes. | Reasoning Graph |
Embedder |
class | Converts text into vector embeddings. | RAG (Retrieval-Augmented Generation) |
EmbeddingProvider |
class | Protocol for embedding providers. | Semantic Cache |
EnvVarNotFoundError |
exception | Raised when a required environment variable is not found. | Configuration Exceptions |
ErrorEvent |
class | An error occurred during processing. | Streaming with Tool Visibility |
EventBusSink |
class | Bridge events to the runtime's EventBus for inter-process notifications. | Agent Events & Notifications |
EventNotifier |
class | Central event coordinator that routes events to configured sinks. | Agent Events & Notifications |
EventSink |
class | Protocol for event notification sinks. | Agent Events & Notifications |
ExportFormat |
enum | Available backends for receiving observability events. | Observability API Reference |
FailureCategory |
enum | Classification of a tool failure. | Adaptive Strategy |
FailureLog |
class | A single tool failure record. | Adaptive Strategy |
FallbackChain |
class | Chain of LLM models with automatic failover. | Model Fallback |
generate_session_id |
function | Generate a cryptographically secure session identifier. | Conversation Persistence |
get_current_caller |
function | Return the CallerContext for the current invocation. |
Building Agents |
GraphState |
class | The complete state of a graph execution. | Reasoning Graph |
GuardrailViolation |
exception | Raised when input or output is blocked by guardrails. | Security Guardrails |
HealthConfig |
class | Behavioral health monitoring configuration. | Runtime Configuration |
HTTPServerConfig |
class | HTTP/SSE MCP server configuration. | SuperAgent Files |
HTTPServerSpec |
class | Specification for a remote MCP server reachable via HTTP/SSE. | MCP Server Configuration for Agents |
IdentityError |
exception | Base exception for every Agent Identity failure. | Agent Identity Security |
IndexReport |
class | Summary of an index operation. |
RAG (Retrieval-Augmented Generation) |
InjectionDetector |
class | Detect prompt injection attacks using a local DeBERTa model. | Security Guardrails |
InMemoryConversationStore |
class | Dict-backed conversation store for testing and prototyping. | Conversation Persistence |
InMemoryProvider |
class | Keyword-search memory provider for testing. | Agent Memory |
InMemoryVectorStore |
class | Reference vector store that keeps everything in memory. | RAG (Retrieval-Augmented Generation) |
load_superagent_file |
function | Convenience function to load a .superagent file with all resolutions. | SuperAgent Files |
LocalEmbeddingProvider |
class | Local embedding via sentence-transformers. | Semantic Cache |
LogSink |
class | Deliver events to Python's logging system. | Agent Events & Notifications |
MCPClient |
class | Production-grade MCP client for a single server. | MCP Client |
MCPClientError |
exception | Raised when an MCP client operation fails. | MCP Client |
MCPConnectionRejectedError |
exception | Raised when an HTTP MCP server rejects the session with a 4xx status. | MCP Client |
MCPCredentialError |
exception | Raised when bearer_token_provider cannot supply a credential. |
MCP Client |
MCPMultiClient |
class | Connect to multiple MCP servers and aggregate their tools. | MCP Client |
MCPToolAdapter |
class | Discover MCP tools and convert them to LangChain BaseTool instances. |
MCP Tool Adapter |
Mem0Provider |
class | Adapter for mem0 <https://github.com/mem0ai/mem0>_. |
Agent Memory |
MemoryAgent |
class | Wraps an agent with automatic memory context injection. | Agent Memory |
MemoryProvider |
class | Interface for memory backends. | Agent Memory |
MemoryResult |
class | A single memory search result. | Agent Memory |
MemorySection |
class | Memory configuration section for .superagent files. | SuperAgent Files |
MissionConfig |
class | Mission-oriented process configuration. | Runtime Configuration |
MissionState |
enum | Mission lifecycle states. | Mission-Oriented Process Model |
Model |
class | Provider, model and credentials in code: the LangChain shape. | Models & Providers |
ModelConfig |
constant | Represent a PEP 604 union type | SuperAgent Files |
ModelSetupError |
exception | A model cannot be used yet: the message says exactly what to do. | Models & Providers |
NERDetector |
class | Detect unstructured PII using GLiNER zero-shot NER model. | Security Guardrails |
NodeResult |
class | Everything that happened during a single node execution. | Reasoning Graph |
ObservabilityConfig |
class | Configuration for the observability system. | Agent Observability |
ObserveLevel |
enum | Controls how much detail the observability system captures. | Agent Observability |
OpenAIEmbeddingProvider |
class | Embedding via OpenAI or Azure OpenAI API. | Semantic Cache |
OptimizationLevel |
enum | Preset optimization levels. | Tool Optimization |
PIICategory |
enum | PII detection categories. Pass a set of these to PromptiseSecurityScanner(enable_pii={...}) to enable only specific PII types. |
Security Guardrails |
PIIDetector |
class | Detect PII using regex patterns with Luhn validation for credit cards. | Security Guardrails |
PostgresConversationStore |
class | PostgreSQL-backed conversation store using asyncpg. |
Conversation Persistence |
ProcessConfig |
class | Configuration for a single agent process. | Runtime Configuration |
ProcessState |
enum | Agent process lifecycle states. | Process Lifecycle |
prompt |
function | Decorator that turns a function into a Prompt. |
Prompt Engineering Overview |
Prompt |
class | Dual-mode prompt: standalone LLM caller + agent instruction source. | Prompt Engineering Overview |
PromptBuilder |
class | Fluent builder for constructing Prompt instances at runtime. |
PromptBuilder |
PromptContext |
class | The agent's complete world during prompt execution. | Prompt Context System |
PromptGraph |
class | A directed graph of reasoning nodes. | Reasoning Graph |
PromptGraphEngine |
class | Adaptive graph traversal engine. | Reasoning Graph |
PromptiseAgent |
class | The unified Promptise agent. | Building Agents |
PromptiseCallbackHandler |
class | LangChain callback handler → ObservabilityCollector bridge. | Callback Handler |
PromptiseSecurityScanner |
class | Unified security scanner for agent input and output. | Security Guardrails |
PromptNode |
class | A complete reasoning unit in the graph. | Reasoning Graph Nodes |
PromptSuite |
class | Group of prompts sharing world configuration. | Prompt Suite & Registry |
PROVIDERS |
constant | tuple of 19 values | Models & Providers |
QueueApprovalHandler |
class | Approval handler using async queues for in-process UIs. | Human-in-the-Loop Approval |
rag_to_tool |
function | Wrap a RAGPipeline as a LangChain tool the agent can call. |
RAG (Retrieval-Augmented Generation) |
RAGPipeline |
class | Orchestrates the full RAG flow: load → chunk → embed → store → retrieve. | RAG (Retrieval-Augmented Generation) |
RecursiveTextChunker |
class | Split text into fixed-size chunks with overlap. | RAG (Retrieval-Augmented Generation) |
RedisConversationStore |
class | Redis-backed conversation store using redis.asyncio. |
Conversation Persistence |
resolve_model |
function | Turn a "provider:model" string (plus optional words) into a chat model. |
Models & Providers |
RetrievalResult |
class | A single hit returned by search. |
RAG (Retrieval-Augmented Generation) |
RuntimeConfig |
class | Top-level runtime configuration (composite). | Runtime Configuration |
SandboxConfigSection |
class | Sandbox configuration section for .superagent files. | SuperAgent Files |
sanitize_memory_content |
function | Sanitize memory content before injection into agent messages. | Agent Memory |
ScanReport |
class | Complete scan result with all findings and metadata. | Security Guardrails |
SecretScopeConfig |
class | Per-process secret scoping configuration. | Runtime Configuration |
SecurityFinding |
class | A single detection result from a scanner. | Security Guardrails |
SemanticCache |
class | Semantic cache for agent responses. | Semantic Cache |
ServerConfig |
callable | Runtime representation of an annotated type. | SuperAgent Files |
servers_to_mcp_config |
function | Convert programmatic server specs to a configuration dict. | Config & Server Specs |
ServerSpec |
constant | Represent a PEP 604 union type | Config & Server Specs |
SessionAccessDenied |
exception | Raised when a user attempts to access a session they do not own. | Conversation Persistence |
SessionInfo |
class | Metadata about a conversation session. | Conversation Persistence |
Severity |
enum | Severity level of a security finding. | Security Guardrails |
SQLiteConversationStore |
class | SQLite-backed conversation store using aiosqlite. |
Conversation Persistence |
StdioServerConfig |
class | Stdio (local process) MCP server configuration. | SuperAgent Files |
StdioServerSpec |
class | Specification for a local MCP server launched via stdio. | MCP Server Configuration for Agents |
StreamEvent |
class | Base event yielded by astream_with_tools(). |
Streaming with Tool Visibility |
SuperAgentConfig |
class | Processed configuration ready for agent building. | SuperAgent Files |
SuperAgentError |
exception | Base exception for all SuperAgent-related errors. | Configuration Exceptions |
SuperAgentLoader |
class | Load and parse .superagent configuration files. | SuperAgent Files |
SuperAgentSchema |
class | Root schema for .superagent YAML files. | SuperAgent Files |
SuperAgentValidationError |
exception | Raised when .superagent schema validation fails. | Configuration Exceptions |
TokenEvent |
class | An LLM token has been generated. | Streaming with Tool Visibility |
tool |
function | Make a Python function a tool an agent can call. | Tools |
ToolEndEvent |
class | A tool has finished executing. | Streaming with Tool Visibility |
ToolError |
exception | Error raised inside a tool handler. | MCP Server Resilience Patterns |
ToolInfo |
class | Human-friendly metadata for a discovered MCP tool. | Tools |
ToolOptimizationConfig |
class | Configuration for MCP tool token optimization. | Tool Optimization |
ToolStartEvent |
class | A tool has started executing. | Streaming with Tool Visibility |
TransporterType |
enum | Available backends for receiving observability events. | Agent Observability |
VectorStore |
class | Persists embedded chunks and performs similarity search. | RAG (Retrieval-Augmented Generation) |
verify_event_signature |
function | Check the X-Promptise-Signature header of a webhook delivery. |
Agent Events & Notifications |
verify_webhook_signature |
function | Check the X-Promptise-Signature of a webhook approval request. |
Human-in-the-Loop Approval |
WebhookApprovalHandler |
class | Approval handler that POSTs to a webhook URL and polls for a decision. | Human-in-the-Loop Approval |
WebhookSink |
class | Deliver events via HTTP POST to a webhook URL. | Agent Events & Notifications |
promptise.engine¶
Reasoning graph. from promptise.engine import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
AutonomousNode |
class | Meta-node: the LLM receives a pool of configured nodes and dynamically composes its own reasoning path at runtime. | Reasoning Graph Nodes |
BudgetHook |
class | Enforces a token and/or cost budget for each graph run. | Graph Hooks & Observability |
CodeActionBudgetError |
exception | A code-action run could not do the whole job within max_tool_calls. |
Reasoning Graph |
CodeActionConfig |
class | Tuning knobs for the code-action pattern. |
Code-Action: one program beats a 20-call tool loop |
CodeActionNode |
class | A node that solves a task by writing and running one sandboxed program. | Reasoning Graph Nodes |
ContextCompaction |
class | Settings for compacting a long tool loop. | Context Lifecycle Management |
CritiqueNode |
class | Challenges the current answer/plan with counter-arguments. | Reasoning Graph Nodes |
CycleDetectionHook |
class | Detect infinite loops by tracking visit patterns. | Graph Hooks & Observability |
EdgeCondition |
class | A built-in edge condition, described by data instead of a lambda. | Graph Serialization & YAML |
ExecutionReport |
class | Summary of a complete graph execution. | Reasoning Graph |
FanOutNode |
class | Sends different sub-questions to parallel child nodes. | Reasoning Graph Nodes |
GraphBudgetExceededError |
exception | A BudgetHook stopped a run over its token or cost budget before the model gave an answer. |
Reasoning Graph |
GraphExecutionError |
exception | A graph run ended on a node failure the graph did not recover from. | Reasoning Graph |
GraphIterationLimitError |
exception | A run hit an iteration limit in the middle of a tool loop. | Reasoning Graph |
GraphMutation |
class | A single graph modification requested during execution. | Reasoning Graph |
GraphStoppedError |
exception | A hook ended a run in the middle of a tool loop. | Reasoning Graph |
GuardNode |
class | Validate state against guards and route based on pass/fail. | Reasoning Graph Nodes |
Hook |
class | Protocol for engine hooks. | Graph Hooks & Observability |
HumanNode |
class | Pause execution and wait for human input. | Reasoning Graph Nodes |
JustifyNode |
class | Forces explicit justification of the last decision/action. | Reasoning Graph Nodes |
LoggingHook |
class | Log every node execution with timing and token counts. | Graph Hooks & Observability |
LoopNode |
class | Repeat a subgraph until a condition is met. | Reasoning Graph Nodes |
MetricsHook |
class | Collects per-node metrics: tokens, latency, errors, call counts. | Graph Hooks & Observability |
node |
function | Decorator that turns an async function into a graph node. | Reasoning Graph Nodes |
NodeEvent |
class | An event emitted during streaming execution. | Reasoning Graph |
NodeFlag |
enum | Typed flags that declare a node's role and capabilities. | Reasoning Graph Nodes |
NodeProtocol |
class | Minimal protocol that every graph node must satisfy. | Reasoning Graph Nodes |
ObserveNode |
class | Processes raw tool results into structured, actionable data. | Reasoning Graph Nodes |
ParallelNode |
class | Run multiple child nodes concurrently and merge results. | Reasoning Graph Nodes |
parse_frontmatter |
function | Extract a YAML-ish frontmatter dict from a Python source file. | Path-Scoped Skill Activation |
PlanNode |
class | Creates structured plans with self-evaluation and subgoal tracking. | Reasoning Graph Nodes |
ReflectNode |
class | Reviews what happened, identifies mistakes, generates corrections. | Reasoning Graph Nodes |
RetryNode |
class | Wraps another node with retry logic and error enrichment. | Reasoning Graph Nodes |
RouterNode |
class | Lightweight LLM call that decides which path to take. | Reasoning Graph Nodes |
Skill |
class | A registered skill with optional path-scoped activation. | Path-Scoped Skill Activation |
SkillRegistry |
class | In-process registry of Skill objects. |
Path-Scoped Skill Activation |
SubgraphNode |
class | Embed a complete sub-graph as a single node. | Reasoning Graph Nodes |
SynthesizeNode |
class | Combines all gathered data into a comprehensive final answer. | Reasoning Graph Nodes |
ThinkNode |
class | Analyzes current state, identifies gaps, recommends next step. | Reasoning Graph Nodes |
TimingHook |
class | Enforce per-node time budgets. | Graph Hooks & Observability |
ToolNode |
class | Execute a specific tool or let the engine pick from available tools. | Reasoning Graph Nodes |
TransformNode |
class | Transform state data without calling the LLM. | Reasoning Graph Nodes |
ValidateNode |
class | LLM evaluates whether output meets specified criteria. | Reasoning Graph Nodes |
promptise.mcp.server¶
MCP server SDK. from promptise.mcp.server import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
APIKeyAuth |
class | API key-based authentication provider. | MCP Server Authentication & Security |
ApprovalDeniedError |
exception | Raised when a server-side approval gate denies (or times out on) a call. | Approval Gates (Server-Side HITL) |
ApprovalGateMiddleware |
class | Enforces human approval on tools declared with requires_approval=True. |
Approval Gates (Server-Side HITL) |
AsymmetricJWTAuth |
class | JWT authentication using asymmetric algorithms (RS256, ES256). | MCP Server Authentication & Security |
AuditIssue |
class | One finding of verify_audit_log. |
MCP Server Observability & Monitoring |
AuditMiddleware |
class | HMAC-chained audit log middleware. | MCP Server Observability & Monitoring |
AuditVerification |
class | Result of verify_audit_log. |
MCP Server Observability & Monitoring |
AuthenticationError |
exception | Raised when authentication fails. | MCP Server Authentication & Security |
AuthMiddleware |
class | Middleware that enforces authentication on tools with auth=True. |
MCP Server Authentication & Security |
AuthProvider |
class | Protocol for authentication providers. | MCP Server Authentication & Security |
BackgroundTasks |
class | Collect and execute background tasks after the handler returns. | MCP Server Resilience Patterns |
build_manifest |
function | Build a JSON-serialisable manifest from a server's registrations. | Advanced MCP Server Patterns |
build_serve_parser |
function | Build the serve subcommand parser. |
MCP Server Deployment |
CacheBackend |
class | Protocol for cache backends. | MCP Server Caching & Performance |
cached |
function | Decorator that caches tool handler results. | MCP Server Caching & Performance |
CacheMiddleware |
class | Server-wide caching middleware. | MCP Server Caching & Performance |
CacheScope |
callable | MCP Server Caching & Performance | |
CancellationToken |
class | Token for checking and responding to client cancellation requests. | MCP Server Resilience Patterns |
CancelledError |
exception | Raised when a tool call is cancelled by the client. | MCP Server Resilience Patterns |
CircuitBreakerMiddleware |
class | Circuit breaker middleware. | MCP Server Resilience Patterns |
CircuitOpenError |
exception | Raised when a tool call is rejected by an open circuit. | MCP Server Resilience Patterns |
CircuitState |
enum | Circuit breaker states. | MCP Server Resilience Patterns |
clear_request_headers |
function | Clear HTTP request headers after the request completes. | MCP Server Routers & Middleware |
ClientContext |
class | Structured information about the authenticated client. | MCP Server Authentication & Security |
ConcurrencyLimiter |
class | Middleware that limits concurrent in-flight tool executions. | MCP Server Caching & Performance |
ConcurrencyLimitError |
exception | Raised when a tool (or the server) is already running its maximum number of concurrent calls. | MCP Server Caching & Performance |
Content |
constant | Represent a PEP 604 union type | Security Guardrails |
CORSConfig |
class | CORS configuration for HTTP and SSE transports. | MCP Server Deployment |
Dashboard |
class | Full-screen tabbed terminal dashboard with keyboard navigation. | MCP Server Observability & Monitoring |
DashboardMiddleware |
class | Outermost middleware that records request metrics for the dashboard. | MCP Server Observability & Monitoring |
DashboardState |
class | Thread-safe shared state between middleware and dashboard renderer. | MCP Server Observability & Monitoring |
DeclaredRateLimitMiddleware |
class | Enforces per-tool rate limits declared via @server.tool(rate_limit=...). |
MCP Server API Reference |
Depends |
function | Mark a parameter for dependency injection. | MCP Server Fundamentals |
dereference_schema |
function | Inline all $ref / $defs in a JSON Schema. |
MCP Server Fundamentals |
ElicitationApprover |
class | Approves by asking the human behind the calling MCP client. | Approval Gates (Server-Side HITL) |
Elicitor |
class | Request structured input from the user mid-execution. | Advanced MCP Server Patterns |
ExceptionHandlerRegistry |
class | Registry mapping exception types to handler functions. | MCP Server Resilience Patterns |
get_context |
function | Return the current request context. | MCP Server Routers & Middleware |
get_request_headers |
function | Return HTTP request headers for the current async context. | MCP Server Routers & Middleware |
Guard |
class | Protocol for tool access guards. | MCP Server Authentication & Security |
HasAllRoles |
class | Guard that requires all of the given roles. | MCP Server Authentication & Security |
HasAllScopes |
class | Guard that requires all of the given OAuth2 scopes. | MCP Server Authentication & Security |
HasRole |
class | Guard that requires any of the given roles. | MCP Server Authentication & Security |
HasScope |
class | Guard that requires any of the given OAuth2 scopes. | MCP Server Authentication & Security |
HasTenant |
class | Guard that requires the client to belong to one of the given tenants. | MCP Server Authentication & Security |
HealthCheck |
class | Health and readiness probe manager. | MCP Server Resilience Patterns |
hot_reload |
function | Run an MCP server with automatic restart on file changes. | Advanced MCP Server Patterns |
ImageContent |
class | Helper for returning image content from tool handlers. | MCP Server Fundamentals |
InMemoryCache |
class | In-process cache with TTL-based expiry and background cleanup. | MCP Server Caching & Performance |
InMemoryQueueBackend |
class | In-process queue backend using asyncio primitives. | MCP Server Job Queue & Background Jobs |
is_upstream_failure |
function | Default failure classifier for CircuitBreakerMiddleware. |
MCP Server Resilience Patterns |
JwksAuth |
class | JWT authentication against an identity provider's JWKS endpoint. | MCP Server Authentication & Security |
JWTAuth |
class | JWT-based authentication provider. | MCP Server Authentication & Security |
LoggingMiddleware |
class | Log every tool call with timing. | MCP Server Routers & Middleware |
MCPError |
exception | Base class for structured MCP server errors. | MCP Server Fundamentals |
MCPQueue |
class | MCP-native job queue for asynchronous background work. | MCP Server Job Queue & Background Jobs |
MCPRouter |
class | Modular grouping for tools, resources, and prompts. | MCP Server Routers & Middleware |
MCPServer |
class | Production-grade MCP server with decorator-based tool registration. | MCP Server Fundamentals |
MetricsCollector |
class | Collects per-tool metrics: call counts, latency, errors. | MCP Server Observability & Monitoring |
MetricsMiddleware |
class | Middleware that records per-tool call metrics. | MCP Server Observability & Monitoring |
Middleware |
class | Protocol for middleware callables. | MCP Server Routers & Middleware |
MiddlewareChain |
class | Executes a stack of middleware around a handler. | MCP Server Routers & Middleware |
mount |
function | Mount a child server's tools, resources, and prompts into a parent. | Advanced MCP Server Patterns |
NamespaceTransform |
class | Prefix all tool names with a namespace. | Advanced MCP Server Patterns |
OnAuthenticateHook |
callable | MCP Server Authentication & Security | |
OpenAPIProvider |
class | Generate MCP tools from an OpenAPI specification. | Advanced MCP Server Patterns |
OTelMiddleware |
class | OpenTelemetry tracing middleware. | MCP Server Observability & Monitoring |
parse_rate_limit |
function | Parse a declared rate-limit string like "100/min". |
MCP Server API Reference |
PendingApprover |
class | Blocks gated calls in a pending store until a human decides. | Approval Gates (Server-Side HITL) |
PerToolConcurrencyLimiter |
class | Middleware that enforces per-tool concurrency limits. | MCP Server Caching & Performance |
ProgressReporter |
class | Report progress during long-running tool calls. | MCP Server Resilience Patterns |
PrometheusMiddleware |
class | Prometheus metrics middleware for MCP tool calls. | MCP Server Observability & Monitoring |
PromptDef |
class | Internal definition of a registered prompt. | MCP Server API Reference |
PromptError |
exception | Error raised inside a prompt handler. | MCP Server API Reference |
QueueBackend |
class | Protocol for pluggable queue storage backends. | MCP Server Job Queue & Background Jobs |
QueueCaller |
class | Identity a queue operation is performed for. | MCP Server Job Queue & Background Jobs |
RateLimitError |
exception | Raised when rate limit is exceeded. | MCP Server Caching & Performance |
RateLimitMiddleware |
class | Middleware that enforces rate limits. | MCP Server Caching & Performance |
RedisCache |
class | Redis-backed cache implementing the CacheBackend protocol. |
MCP Server Caching & Performance |
register_batch_tool |
function | Register a batch meta-tool on the server. | Advanced MCP Server Patterns |
register_manifest |
function | Register a docs://manifest resource on the server. |
Advanced MCP Server Patterns |
RequestContext |
class | Per-request context available to handlers and middleware. | MCP Server Fundamentals |
RequireAuth |
class | Guard that requires an authenticated client_id on the context. |
MCP Server Authentication & Security |
RequireClientId |
class | Guard that requires a specific client_id. |
MCP Server Authentication & Security |
RequireTenant |
class | Guard that requires the client to belong to some tenant. | MCP Server Authentication & Security |
resolve_server |
function | Import and return the server instance from a target string. | MCP Server Deployment |
ResourceDef |
class | Internal definition of a registered resource or resource template. | MCP Server API Reference |
ResourceError |
exception | Error raised inside a resource handler. | MCP Server API Reference |
run_serve |
function | Execute the serve command. |
MCP Server Deployment |
Sampler |
class | Request LLM completions from the client. | Advanced MCP Server Patterns |
ServerLogger |
class | Send log messages from the server to the connected MCP client. | MCP Server Observability & Monitoring |
ServerSettings |
class | Base server settings with sensible defaults. | MCP Server Fundamentals |
SessionManager |
class | Manages per-session state stores. | Advanced MCP Server Patterns |
SessionState |
class | Key-value state scoped to the current MCP session. | Advanced MCP Server Patterns |
set_request_headers |
function | Store HTTP request headers for the current async context. | MCP Server Routers & Middleware |
StreamingResult |
class | Collects partial results for streaming-style tool responses. | Advanced MCP Server Patterns |
StructuredLoggingMiddleware |
class | Emit structured JSON log entries for every tool call. | MCP Server Observability & Monitoring |
TagFilterTransform |
class | Only expose tools that have at least one of the required tags. | Advanced MCP Server Patterns |
TestClient |
class | MCP Server Testing | |
TimeoutMiddleware |
class | Enforce per-call timeout. | MCP Server Routers & Middleware |
TokenBucketLimiter |
class | Token bucket rate limiter. | MCP Server Caching & Performance |
TokenEndpointConfig |
class | Configuration for the built-in OAuth 2.0 token issuer. | MCP Server Authentication & Security |
ToolAnnotations |
class | MCP tool annotations describing tool behaviour hints. | MCP Server Fundamentals |
ToolDef |
class | Internal definition of a registered tool. | MCP Server Fundamentals |
ToolResponse |
class | Response wrapper that lets handlers return content with metadata. | MCP Server Fundamentals |
ToolTransform |
class | Protocol for tool transforms. | Advanced MCP Server Patterns |
TransportType |
enum | Supported MCP transport protocols. | MCP Server Deployment |
ValidationError |
exception | Raised when input validation fails. | Code Sandbox for Agents |
verify_audit_log |
function | Verify the HMAC chain of an audit log file written by AuditMiddleware. |
MCP Server Observability & Monitoring |
VersionedToolRegistry |
class | Overlay registry that manages versioned tools. | Advanced MCP Server Patterns |
VisibilityTransform |
class | Hide tools based on a predicate. | Advanced MCP Server Patterns |
WebhookMiddleware |
class | Fire webhooks on tool call events. | MCP Server Production Features |
promptise.mcp.client¶
MCP client. from promptise.mcp.client import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
BearerTokenProvider |
callable | MCP Client | |
InFlightToolCall |
class | A call_tool request awaiting its result. |
MCP Client |
MCPToolError |
exception | An MCP tool ran and reported a failure. | MCP Tool Adapter |
promptise.mcpcast¶
MCPcast. from promptise.mcpcast import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
api_name_from_spec |
function | A slug for the API, from info.title. |
MCPcast an Existing API |
ApiPlan |
class | The upstream API and how the generated server talks to it. | MCPcast a Real API: Recipes |
apply_curation |
function | Turn an accepted proposal into a validated MCPcastPlan. |
MCPcast API Reference |
ApprovalMode |
enum | Who approves approval-gated tool calls. | MCPcast API Reference |
AuthMode |
enum | How the generated server authenticates against the upstream API. | MCPcast an Existing API |
build_plan |
function | Map operations 1:1 onto tools, filtered by profile. | MCPcast an Existing API |
CallRecorder |
class | Records every tool call the agent makes, grouped by task id. | Lab: MCPcast Your SaaS API |
check_document |
function | Refuse whole-document defects before anything walks or copies document. | MCPcast API Reference |
check_postconditions |
function | Every post-condition violation in result (empty means it passed). | MCPcast an Existing API |
Classification |
class | The outcome of classifying one operation, with its reasoning. | Adaptive Strategy |
classify |
function | Classify op and explain the decision. | MCPcast an Existing API |
classify_operation |
function | The RiskClass for op (see classify for the reasoning). |
MCPcast API Reference |
Completer |
callable | MCPcast an Existing API | |
ConfusedPair |
class | The agent chose chosen first when expected was the target. | MCPcast API Reference |
credential_slot |
function | The one slot the plan's server presents its credential in, from the spec's schemes. | MCPcast API Reference |
CredentialSlot |
class | Where the generated server presents the upstream credential. | MCPcast API Reference |
curate |
function | Design the tool surface with a model, enforcing every post-condition in code. | MCPcast an Existing API |
CuratedParam |
class | The model's adjustments to one parameter. | MCPcast API Reference |
CuratedTool |
class | One tool as proposed by the model. | MCPcast API Reference |
CurationResult |
class | The model's complete proposal. | MCPcast API Reference |
CurationViolation |
exception | A curation proposal broke one or more post-conditions. | MCPcast API Reference |
derive_tool_name |
function | getPetById → get_pet_by_id (unique within taken, if given). |
MCPcast API Reference |
describe_written |
function | "bookshelf_mcp/ (7 modules), tests/ (2), server.py, README.md, pyproject.toml…". |
MCPcast API Reference |
DroppedOp |
class | An operation deliberately left out of the tool surface. | MCPcast API Reference |
EvalReport |
class | The Agent Readiness Score with its evidence. | MCPcast API Reference |
EvalTask |
class | One user request and the tool it is designed to exercise. | MCPcast an Existing API |
evaluate |
function | Run the Agent Readiness evaluation. | MCPcast an Existing API |
example_mismatch |
function | Why value does not fit schema, or None if it plausibly does. |
MCPcast API Reference |
example_value |
function | A plausible example for one JSON Schema (used for tool examples and mocks). | MCPcast API Reference |
expanded_nodes |
function | Count the nodes of document as extraction would visit them, stopping past limit. | MCPcast API Reference |
extract_operations |
function | Extract every operation from spec, in document order. | MCPcast an Existing API |
generate_tasks |
function | Generate up to count tasks from the plan's tool set. | MCPcast API Reference |
grade_for |
function | Letter grade for a score in [0, 1]. |
MCPcast API Reference |
is_plan_document |
function | True if a parsed YAML/JSON document looks like an MCPcastPlan rather than an OpenAPI spec. |
MCPcast API Reference |
is_url |
function | True if source is an http(s):// URL string (the scheme is case-insensitive). |
MCPcast API Reference |
load_generated_server |
function | Import a generated project's server.py launcher as a module. |
MCPcast an Existing API |
load_spec |
function | Load an OpenAPI document from a URL, file path, raw JSON/YAML text, or dict. | MCPcast an Existing API |
make_example |
function | One worked example covering every required, visible parameter. | MCPcast API Reference |
MAX_DOCUMENT_NODES |
constant | MCPcast API Reference | |
mcpcast |
function | The deterministic pipeline: load → extract → classify → plan. | MCPcast an Existing API |
MCPcastError |
exception | Raised for an invalid spec, plan, or curation result. | MCPcast an Existing API |
MCPcastPlan |
class | The complete, validated tool plan for one API. | MCPcast an Existing API |
mock_transport |
function | An httpx transport: live GET/HEAD (optional), spec-derived mocks otherwise. |
MCPcast an Existing API |
Operation |
class | One HTTP operation extracted from the spec. | MCPcast an Existing API |
package_name |
function | The import package for an API name: bookshelf → bookshelf_mcp. |
MCPcast API Reference |
ParamPlan |
class | The agent-facing shape of one tool parameter. | Lab: MCPcast Your SaaS API |
ParamSpec |
class | One input of an operation, with its wire location. | MCPcast API Reference |
plain_http_hosts |
function | Hosts the generated server would send its credential to over plain http://. |
MCPcast an Existing API |
public_url |
function | url with its userinfo, query string and fragment removed. | MCPcast API Reference |
render_curation_prompt |
function | The user prompt: the operation catalogue plus the budget. | MCPcast API Reference |
render_project |
function | Every generated file for plan, keyed by path relative to the project root. | MCPcast an Existing API |
render_readme |
function | Render README.md for plan. |
MCPcast an Existing API |
RESERVED_TOOL_NAMES |
constant | frozenset of 18 values | MCPcast API Reference |
resolve_base_url |
function | The plan's API base URL: override, else the document-level server, else the first operation's own base. | MCPcast API Reference |
risk_floor |
function | The lowest class classify could assign to an operation with this wire mapping. |
MCPcast API Reference |
RiskClass |
enum | What calling a tool can do to the world. | MCPcast an Existing API |
RouteParam |
class | Where one parameter of a route travels on the wire. | MCPcast API Reference |
RoutePlan |
class | One upstream HTTP operation a tool can execute. | MCPcast API Reference |
SafetyProfile |
enum | How much of the API an agent is allowed to reach. | MCPcast an Existing API |
SCAFFOLD_ONCE |
constant | frozenset of 4 values | MCPcast API Reference |
score |
function | Compute the report from task results (pure; no model, no network). | MCPcast an Existing API |
scrub_strings |
function | document with every string: keys and values, at any depth: passed through scrub_text. |
MCPcast API Reference |
SecurityScheme |
class | One components.securitySchemes entry (Swagger 2: securityDefinitions), resolved. |
MCPcast API Reference |
spec_base_url |
function | The API base URL: override, else servers[0].url, else Swagger 2 host. |
MCPcast API Reference |
spec_description |
function | The spec's info.description, stripped (empty if absent or if info is not a mapping). |
MCPcast API Reference |
spec_summary_line |
function | One line for the plan's api.description (and the server's instructions): the title, plus the first line of the description when it adds something (`"Bookshelf API: Books,... |
MCPcast API Reference |
spec_title |
function | The spec's info.title, stripped (empty if absent or if info is not a mapping). |
MCPcast API Reference |
TaskResult |
class | What happened when the agent attempted one task. | MCPcast API Reference |
tool_group |
function | The tools module a tool lives in: its first tag, else the resource in its path. | MCPcast API Reference |
ToolCall |
class | One tool invocation observed during a task. | MCPcast API Reference |
ToolPlan |
class | One generated MCP tool. | Lab: MCPcast Your SaaS API |
tools_from_server |
function | LangChain tools that call server in-process through TestClient. |
Lab: MCPcast Your SaaS API |
unmappable_reason |
function | Why op cannot become a tool, or None if it can. |
MCPcast API Reference |
valid_tool_name |
function | Why name cannot be a tool name, or None if it can. |
MCPcast API Reference |
write_eval |
function | Write eval/tasks.yaml and eval/report.md under out_dir. |
MCPcast an Existing API |
write_project |
function | Write the project for plan into out_dir. | MCPcast an Existing API |
promptise.prompts¶
Prompting. from promptise.prompts import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
advisor |
constant | Balanced advisory perspective. | Prompt Strategies & Perspectives |
analyst |
constant | Evidence-based analysis perspective. | Prompt Strategies & Perspectives |
AssembledPrompt |
class | Result of assembling prompt blocks. | Layer 1: PromptBlocks |
Block |
class | A composable unit of prompt content. | Layer 1: PromptBlocks |
block |
function | Decorator that turns a function into a Block. | Layer 1: PromptBlocks |
BlockContext |
class | Runtime context available to blocks during rendering. | Layer 1: PromptBlocks |
blocks |
function | Decorator that attaches blocks to a Prompt. |
Layer 1: PromptBlocks |
branch |
function | Route to different prompts based on a condition. | Prompt Chaining |
chain |
function | Create a sequential chain of prompts. | Prompt Chaining |
chain_of_thought |
constant | Step-by-step reasoning with an answer marker. | Prompt Strategies & Perspectives |
Composite |
class | Groups multiple blocks as a single unit. | Layer 1: PromptBlocks |
Conditional |
class | Block that renders only when a condition is true. | Layer 1: PromptBlocks |
constraint |
function | Decorator that attaches a constraint to a Prompt. |
Prompt Engineering Overview |
context |
function | Decorator that attaches context providers to a Prompt. |
Prompt Context System |
ContextProvider |
class | Pluggable source of dynamic context for prompts. | Prompt Context System |
ContextSlot |
class | Dynamic injection point, filled at runtime (priority configurable). | Layer 1: PromptBlocks |
ConversationContext |
class | Conversation history and state. | Prompt Context System |
ConversationFlow |
class | Base class for conversation flow state machines. | Layer 2: ConversationFlow |
creative |
constant | Creative exploration perspective. | Prompt Strategies & Perspectives |
critic |
constant | Critical evaluation perspective. | Prompt Strategies & Perspectives |
decompose |
constant | Decompose into subproblems, solve each, then combine. | Prompt Strategies & Perspectives |
EnvironmentContext |
class | Runtime environment. Extend with deployment-specific fields. | Prompt Context System |
ErrorContext |
class | Previous errors for retry and recovery. | Prompt Context System |
Examples |
class | Few-shot examples with auto-truncation (priority 4). | Layer 1: PromptBlocks |
fallback |
function | Try prompts in order until one succeeds. | Prompt Chaining |
Guard |
class | Protocol for prompt input/output guards. | Prompt Guards |
guard |
function | Decorator that attaches guards to a Prompt. |
Prompt Guards |
GuardError |
exception | Raised when a guard rejects input or output. | Prompt Guards |
Identity |
class | Who the agent is: always included (priority 10). | Layer 1: PromptBlocks |
load_directory |
function | Load all .prompt files from a directory into a registry. |
Prompt Loader & Templates |
load_prompt |
function | Load a prompt from a .prompt YAML file. |
Prompt Loader & Templates |
load_url |
function | Load a prompt from a URL (e.g. GitHub raw file). | Prompt Loader & Templates |
ObservationBlock |
class | Inject recent tool results into the prompt. | Layer 1: PromptBlocks |
OutputContext |
class | Expected output characteristics. | Prompt Context System |
OutputFormat |
class | Response structure specification (priority 8). | Layer 1: PromptBlocks |
parallel |
function | Execute multiple prompts concurrently. | Prompt Chaining |
Perspective |
class | Protocol for cognitive perspectives. | Prompt Strategies & Perspectives |
perspective |
function | Create a CustomPerspective. |
Prompt Strategies & Perspectives |
Phase |
class | A named phase in a conversation flow. | Layer 2: ConversationFlow |
phase |
function | Decorator that marks a method as a phase handler. | Layer 2: ConversationFlow |
PhaseBlock |
class | Stage-specific instructions that change per reasoning phase. | Layer 1: PromptBlocks |
plan_and_execute |
constant | Plan first, then execute each step. | Prompt Strategies & Perspectives |
PlanBlock |
class | Render the current plan with subgoal progress. | Layer 1: PromptBlocks |
PromptAssembler |
class | Assembles blocks into a final prompt with optional token budgeting. | Layer 1: PromptBlocks |
PromptFileError |
exception | Error loading or saving a .prompt file. |
Prompt Loader & Templates |
PromptInspector |
class | Collects and displays prompt assembly and execution traces. | Prompt Inspector & Observability |
PromptObserver |
class | Bridge between prompt execution and ObservabilityCollector. | Prompt Inspector & Observability |
PromptRegistry |
class | Singleton registry for versioned prompts. | Prompt Suite & Registry |
PromptStats |
class | Execution statistics for a single prompt call. | Prompt Engineering Overview |
PromptTestCase |
class | Base class for prompt tests. | Prompt Testing |
PromptTrace |
class | Complete trace of a prompt assembly + execution. | Prompt Inspector & Observability |
PromptValidationError |
exception | Schema validation failed for a .prompt file. |
Prompt Loader & Templates |
ReflectionBlock |
class | Inject past learnings from reflection/evaluation. | Layer 1: PromptBlocks |
registry |
constant | Singleton registry for versioned prompts. | Prompt Suite & Registry |
render_template |
function | Convenience function: create a TemplateEngine and render. |
Prompt Loader & Templates |
retry |
function | Wrap a prompt with exponential backoff retry. | Prompt Chaining |
Rules |
class | Behavioral constraints (priority 9). | Layer 1: PromptBlocks |
save_prompt |
function | Save a Prompt to a YAML file. |
Prompt Loader & Templates |
Section |
class | Custom named section (priority configurable). | Layer 1: PromptBlocks |
self_critique |
constant | Generate, critique, and improve. | Prompt Strategies & Perspectives |
ShellExecutionError |
exception | Raised when a templated shell command fails or times out. | Shell Context Injection in Templates |
ShellExecutor |
class | Protocol for objects that can run a templated shell command. | Shell Context Injection in Templates |
SimpleBlock |
class | A block created from a string or callable. | Layer 1: PromptBlocks |
Strategy |
class | Protocol for reasoning strategies. | Prompt Strategies & Perspectives |
structured_reasoning |
constant | Multi-step structured reasoning with customizable phases. | Prompt Strategies & Perspectives |
SubprocessShellExecutor |
class | Default shell executor backed by run. |
Shell Context Injection in Templates |
TeamContext |
class | Other agents in the team. | Prompt Context System |
TemplateEngine |
class | Template engine with include registry and optional shell interpolation. | Prompt Loader & Templates |
ToolsBlock |
class | Render available tool schemas for the LLM. | Layer 1: PromptBlocks |
TurnContext |
class | Mutable context passed to @phase handlers each turn. |
Layer 2: ConversationFlow |
UserContext |
class | Who the agent is serving. Extend with any user-specific fields. | Prompt Context System |
version |
function | Decorator that registers a Prompt in the global registry. |
Prompt Suite & Registry |
promptise.runtime¶
Agent Runtime. from promptise.runtime import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
AgentManifestSchema |
class | Root schema for .agent manifest files. |
Agent Manifests |
Anomaly |
class | A detected behavioral anomaly. | Behavioral Health Monitoring |
AnomalyType |
enum | Types of behavioral anomaly. | Behavioral Health Monitoring |
BaseTrigger |
class | Protocol for all trigger implementations. | Triggers Overview |
BLOCKABLE_EVENTS |
constant | frozenset of 6 values | Runtime Lifecycle Hooks |
BudgetEnforcer |
class | Handles budget violations according to the configured policy. | Autonomy Budget |
BudgetExceededError |
exception | Raised before a tool call that would exceed an autonomy budget limit. | Agent Runtime |
BudgetState |
class | Thread-safe budget counters with per-run and daily tracking. | Autonomy Budget |
BudgetViolation |
class | Describes a budget limit that was exceeded. | Autonomy Budget |
BusEvent |
class | An event delivered by InProcessEventBus. |
Event and Webhook Triggers |
ClusterNode |
class | This runtime's membership in a cluster (see the module docstring). | Service Discovery and Transport |
ContextConfig |
class | AgentContext configuration. | Runtime Configuration |
ConversationBuffer |
class | Rolling buffer of conversation messages (short-term memory). | Conversation Buffer |
create_trigger |
function | Factory: create a trigger from a TriggerConfig. |
Triggers Overview |
CronTrigger |
class | Fires at scheduled intervals defined by a cron expression. | Cron Trigger |
DEFAULT_DEVELOPMENT_CONFIG |
constant | Top-level runtime configuration (composite). | Runtime Configuration |
DEFAULT_PRODUCTION_CONFIG |
constant | Top-level runtime configuration (composite). | Runtime Configuration |
DiscoveredNode |
class | A discovered runtime node. | Service Discovery and Transport |
DispatchResult |
class | Outcome of a dispatch call. |
Runtime Lifecycle Hooks |
DistributedConfig |
class | Distributed runtime coordination (RuntimeConfig.distributed). |
Runtime Configuration |
EscalationTarget |
class | Where to send escalation notifications. | Runtime Configuration |
EventTrigger |
class | Trigger that fires when an EventBus event matches. | Triggers Overview |
ExecutionMode |
enum | Agent execution mode. | Runtime Configuration |
FileJournal |
class | JSONL-based journal for durable process audit logging. | Journal Backends |
HealthMonitor |
class | Lightweight behavioral health analysis. | Behavioral Health Monitoring |
HookBlocked |
exception | Raised by a hook to block the action that triggered it. | Runtime Lifecycle Hooks |
HookCallable |
callable | Runtime Lifecycle Hooks | |
HookConfig |
class | One runtime lifecycle hook of a process (ProcessConfig.hooks). |
Runtime Configuration |
HookContext |
class | Context passed to every hook callback. | Runtime Lifecycle Hooks |
HookEvent |
enum | Lifecycle events that a HookManager can dispatch. |
Runtime Lifecycle Hooks |
HookManager |
class | Event-keyed dispatcher for runtime lifecycle hooks. | Runtime Lifecycle Hooks |
InboxConfig |
class | Configuration for the human-to-agent message inbox. | Runtime Configuration |
InboxMessage |
class | A message from a human to a running agent. | Orchestration API |
InboxResponse |
class | An agent's response to a question. | Orchestration API |
InMemoryJournal |
class | In-memory journal for testing and development. | Journal Backends |
InProcessEventBus |
class | An event bus that delivers to subscribers in the same Python process. | Event and Webhook Triggers |
JournalConfig |
class | Journal (durable audit log) configuration. | Runtime Configuration |
JournalEntry |
class | Single journal record. | Journal Backends |
JournalError |
exception | Raised when journal read, write, or replay fails. | Agent Runtime |
JournalLevel |
enum | Journal detail level. | Journal System |
JournalProvider |
class | Protocol for journal backends. | Journal Backends |
load_manifest |
function | Load and validate a .agent manifest file. |
Agent Manifests |
manifest_to_process_config |
function | Convert a manifest into a ProcessConfig. |
Agent Manifests |
ManifestEnvVarError |
exception | A ${VAR} in a .agent manifest names an unset environment variable. |
Agent Runtime |
ManifestError |
exception | Raised when .agent manifest loading fails (I/O or parse error). |
Agent Runtime |
ManifestValidationError |
exception | Raised when .agent manifest schema validation fails. |
Agent Runtime |
MessageInbox |
class | Async-safe inbox for human-to-agent messages. | Orchestration API |
MessageTrigger |
class | Trigger that fires when a message arrives on a broker topic. | Triggers Overview |
MessageType |
enum | Types of human-to-agent messages. | Orchestration API |
MissionEvaluation |
class | Result of an evaluation. | Mission-Oriented Process Model |
MissionEvidence |
class | Bundle of evidence passed to the evaluator. | Mission-Oriented Process Model |
MissionTracker |
class | Tracks mission progress across invocations. | Mission-Oriented Process Model |
NodeInfo |
class | Information about a runtime node in the cluster. | Distributed Coordinator |
OpenModeConfig |
class | Guardrails for open execution mode. | Runtime Configuration |
OrchestrationAPI |
class | REST API server for managing an AgentRuntime. | Orchestration API |
OrchestrationAPIError |
exception | An error answer from the Orchestration API. | Orchestration API |
OrchestrationClient |
class | Async client for the Promptise Orchestration API. | Orchestration API |
ProcessLifecycle |
class | State machine for process lifecycle. | Process Lifecycle |
ProcessStateError |
exception | Raised when a lifecycle state transition is invalid. | Agent Runtime |
ProcessTransition |
class | Record of a single state transition. | Process Lifecycle |
register_trigger_type |
function | Register a custom trigger type. | Triggers Overview |
registered_trigger_types |
function | Return the names of all registered trigger types. | Triggers Overview |
RegistryDiscovery |
class | In-process registry for node discovery. | Service Discovery and Transport |
ReplayEngine |
class | Replays journal entries to reconstruct process state. | Replay Engine |
RewindEngine |
class | Multi-granularity rewind over a journal. | Agent Process Rewind |
RewindMode |
enum | Selectivity of a apply operation. |
Agent Process Rewind |
RewindPlan |
class | Result of plan: preview without committing. |
Agent Process Rewind |
RewindResult |
class | State produced by apply. |
Agent Process Rewind |
RuntimeBaseError |
exception | Base exception for all runtime module errors. | Agent Runtime |
RuntimeCoordinator |
class | Cluster coordinator for distributed runtime management. | Distributed Coordinator |
RuntimeDashboard |
class | Full-screen tabbed terminal dashboard for the Agent Runtime. | Runtime Dashboard |
RuntimeDashboardState |
class | Thread-safe shared state for the runtime dashboard. | Runtime Dashboard |
RuntimeDataCollector |
class | Collects real-time data from a running AgentRuntime into DashboardState. | Runtime Dashboard |
save_manifest |
function | Write a manifest to a YAML file. | Agent Manifests |
SecretScope |
class | Per-process secret vault with TTL, access logging, and rotation. | Secret Scoping |
ShellHook |
class | Runtime hook that shells out to an external command. | Runtime Lifecycle Hooks |
ShellHookResult |
class | Parsed result returned by an external shell hook. | Runtime Lifecycle Hooks |
StateEntry |
class | Audit record for a single state write. | Agent Context |
StateError |
exception | Raised when an invalid state transition is attempted. | Process Lifecycle |
StaticDiscovery |
class | Static discovery from a known list of node URLs. | Service Discovery and Transport |
ToolCostAnnotation |
class | Cost and reversibility annotation for a tool. | Runtime Configuration |
TriggerConfig |
class | Configuration for a single trigger. | Runtime Configuration |
TriggerDeliveryConfig |
class | How a process handles trigger events that can't be processed. | Runtime Configuration |
TriggerError |
exception | Raised when a trigger fails to start, stop, or produce events. | Agent Runtime |
TriggerEvent |
class | Event produced by a trigger. | Triggers Overview |
VALID_TRANSITIONS |
constant | mapping of 8 entries | Process Lifecycle |
validate_manifest |
function | Validate a manifest file, returning a list of warnings. | Agent Manifests |
promptise.runtime.triggers¶
Runtime triggers. from promptise.runtime.triggers import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
BaseTrigger |
class | Protocol for all trigger implementations. | Triggers Overview |
BusEvent |
class | An event delivered by InProcessEventBus. |
Event and Webhook Triggers |
compile_filter |
function | Turn a filter_expression into a predicate over trigger events. |
Triggers Overview |
create_trigger |
function | Factory: create a trigger from a TriggerConfig. |
Triggers Overview |
CronTrigger |
class | Fires at scheduled intervals defined by a cron expression. | Cron Trigger |
EventTrigger |
class | Trigger that fires when an EventBus event matches. | Triggers Overview |
FilterExpressionError |
exception | A filter expression is malformed or uses a construct that is not allowed. | Triggers Overview |
InProcessEventBus |
class | An event bus that delivers to subscribers in the same Python process. | Event and Webhook Triggers |
MessageTrigger |
class | Trigger that fires when a message arrives on a broker topic. | Triggers Overview |
register_trigger_type |
function | Register a custom trigger type. | Triggers Overview |
registered_trigger_types |
function | Return the names of all registered trigger types. | Triggers Overview |
TriggerEvent |
class | Event produced by a trigger. | Triggers Overview |
unregister_trigger_type |
function | Remove a registered trigger type. | Triggers Overview |
promptise.runtime.journal¶
Runtime journal. from promptise.runtime.journal import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
FileJournal |
class | JSONL-based journal for durable process audit logging. | Journal Backends |
InMemoryJournal |
class | In-memory journal for testing and development. | Journal Backends |
JournalEntry |
class | Single journal record. | Journal Backends |
JournalLevel |
enum | Journal detail level. | Journal System |
JournalProvider |
class | Protocol for journal backends. | Journal Backends |
RedactingJournal |
class | A JournalProvider that redacts before it writes. |
Journal Backends |
ReplayEngine |
class | Replays journal entries to reconstruct process state. | Replay Engine |
RewindEngine |
class | Multi-granularity rewind over a journal. | Agent Process Rewind |
RewindMode |
enum | Selectivity of a apply operation. |
Agent Process Rewind |
RewindPlan |
class | Result of plan: preview without committing. |
Agent Process Rewind |
RewindResult |
class | State produced by apply. |
Agent Process Rewind |
promptise.runtime.distributed¶
Distributed runtime. from promptise.runtime.distributed import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
ClusterNode |
class | This runtime's membership in a cluster (see the module docstring). | Service Discovery and Transport |
DiscoveredNode |
class | A discovered runtime node. | Service Discovery and Transport |
NodeInfo |
class | Information about a runtime node in the cluster. | Distributed Coordinator |
NodeUnreachableError |
exception | A node could not be reached (connection refused, DNS, timeout...). | Distributed Coordinator |
ProcessDiscovery |
class | Protocol for node discovery mechanisms. | Service Discovery and Transport |
RegistryDiscovery |
class | In-process registry for node discovery. | Service Discovery and Transport |
RuntimeCoordinator |
class | Cluster coordinator for distributed runtime management. | Distributed Coordinator |
RuntimeTransport |
class | HTTP management API for an AgentRuntime. |
Service Discovery and Transport |
StaticDiscovery |
class | Static discovery from a known list of node URLs. | Service Discovery and Transport |
promptise.identity¶
Agent identity. from promptise.identity import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
AwsEksProjectedProvider |
class | AWS EKS projected-token credential source. | AWS IAM |
AwsStsProvider |
class | AWS STS web-identity-token credential source (lazily imports boto3). | AWS IAM |
CachedCredential |
class | One cached identity credential (a verifiable JWT) and its expiry. | Agent Identity Architecture |
CallableTokenProvider |
class | Invokes a user-supplied callable to acquire the identity JWT. | Agent Identity Architecture |
CREDENTIAL_REFRESH_BUFFER_SECONDS |
constant | Agent Identity Architecture | |
CredentialAcquisitionError |
exception | A verifiable credential provider could not produce an identity JWT. | Agent Identity Architecture |
decode_jwt_claims |
function | Best-effort decode of a JWT's claims payload, without verifying. | Agent Identity Architecture |
decode_jwt_expiry |
function | Best-effort read of the exp claim from a JWT, without verifying. |
Agent Identity Architecture |
EntraManagedIdentityProvider |
class | Entra managed-identity credential source backed by Azure IMDS. | Microsoft Entra ID |
EntraProjectedTokenProvider |
class | Entra projected-token credential source for AKS Workload Identity. | Microsoft Entra ID |
FileTokenProvider |
class | Reads the identity JWT from a file on disk on every refresh. | Agent Identity Architecture |
GcpMetadataProvider |
class | GCP credential source backed by the Compute metadata server. | Google Cloud |
IdentityProvider |
class | Abstract base for a verifiable agent-identity credential source. | Agent Identity Architecture |
OidcCallableProvider |
class | Generic OIDC credential source that invokes a callable for the JWT. | Generic OIDC |
OidcFileProvider |
class | Generic OIDC credential source that reads the JWT from a file path. | Generic OIDC |
PlatformDetectionError |
exception | auto could not detect a supported platform. |
Agent Identity Architecture |
ProviderConfigError |
exception | A credential provider was misconfigured at construction time. | Agent Identity Architecture |
SpiffeFileProvider |
class | SPIFFE credential source reading a JWT-SVID from spiffe-helper. |
SPIFFE / SPIRE |
SpiffeSdkProvider |
class | SPIFFE credential source using the Workload API via py-spiffe. | SPIFFE / SPIRE |
promptise.sandbox¶
Sandbox. from promptise.sandbox import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
cleanup_on_exit |
function | Ensure session is cleaned up on process exit. | Sandbox API Reference |
CommandResult |
class | Result of a command execution in the sandbox. | Code Sandbox for Agents |
DockerBackend |
class | Docker-based sandbox backend with gVisor support. | Code Sandbox for Agents |
NetworkMode |
enum | Network isolation modes for sandbox. | Code Sandbox for Agents |
SandboxBackend |
class | Abstract base class for sandbox backends. | Code Sandbox for Agents |
SandboxConfig |
class | Configuration for sandbox environment. | Code Sandbox for Agents |
SandboxContainerManager |
class | Utility for managing and monitoring sandbox containers. | Sandbox API Reference |
SandboxManager |
class | Manages sandbox lifecycle and provides integration with agents. | Code Sandbox for Agents |
SandboxSession |
class | Manages a persistent sandbox session for command execution. | Code Sandbox for Agents |
promptise.run¶
promptise run. from promptise.run import ...
| Name | Kind | Summary | Documented in |
|---|---|---|---|
find_manifest |
function | The nearest manifest from start (default: the cwd) upwards, or None. |
Projects & promptise run |
load_manifest |
function | Read and validate the manifest at path (promptise.toml or pyproject.toml). |
Projects & promptise run |
ManifestError |
exception | The manifest cannot be found, read or validated. | Projects & promptise run |
ProjectManifest |
class | A parsed promptise.toml (or [tool.promptise]). |
Projects & promptise run |
RunContext |
class | What an agent factory may want to know about the run. | Projects & promptise run |