MCP Path
Crawl -> Learn Loop
Crawl -> Learn Loop is a public reference for edge cases, product boundaries, and robust failure paths. It names the signal, policy, or flow an agent should understand before choosing a concrete tool.
Reference page for a documented MCP capability path.
- Type
- MCP path
- Family
- Cross-Feature Integration
- Effect
- bounded run
- Status
- Reference
- Path
- 50.1
Purpose
What this entry explains
What it does
This reference explains Crawl -> Learn Loop for edge cases, product boundaries, and robust failure paths. It is kept as a named reference so agents can cite the flow without inventing a tool name.
Use when
- Use this entry when an agent needs to start or monitor the bounded flow "Crawl -> Learn Loop" for edge cases, product boundaries, and robust failure paths.
- Use it as a reference path when the catalog describes a capability but no single public tool name is explicit.
- Use it before chaining follow-up tools so the next step is based on current evidence.
Reference Use
How agents should cite and apply this area
Examples are maintained at family level and use only public tool names or reference paths already present in the catalog.
Crawl -> Learn Loop describes a gate for edge cases, product boundaries, and robust failure paths. The path shows which signal, gate, behavior, or boundary must be checked before choosing a concrete tool.
An agent cites this path when it needs Crawl -> Learn Loop as context for a decision, block, target check, or follow-up tool choice.
The public source does not name one callable tool for this path. The documentation therefore keeps it as a reference path and does not invent a callable name.
Relevant response signals: perceive, pks_upsert. Safety axes: Read current state, External, Automation, User confirmation. The reference path alone is not permission to execute. Before acting, check current MCP discovery, visible target, scope, and the actual response.
Family example
A task in edge cases, product boundaries, and robust failure paths can trigger powerful execution and therefore needs target, approval, and result check before the step.
The agent starts with Crawl -> Learn Loop, reads the current response or reference, and only then chooses the concrete next tool.
Current discovery, target, user control, warning signals, and result check come before execution.Contract
Inputs and important response fields
This page is a public reference. Agents and integrators should still read current MCP tool discovery before execution, because schemas can be gated by settings or version.
Inputs
No stable public input field is derived from the catalog source for this path. Read current MCP discovery before execution.
| Response field | Explanation |
|---|---|
perceive | Response field named by the catalog source. Treat it as current evidence for the next decision. |
pks_upsert | Response field named by the catalog source. Treat it as current evidence for the next decision. |
Safety
Boundary before execution
Starts or observes a bounded run. Scope, limits, progress, and terminal status need to stay visible.
Set scope and limits before starting the run, poll explicit progress or status fields, and stop instead of assuming completion.
For humans, this entry shows which bounded flow in edge cases, product boundaries, and robust failure paths starts or continues, and where scope, progress, and stopping conditions belong.
High-Impact Review
Execution boundary and recheck hints
Review category: Proxy/external connection
External targets, proxies, gateways, and transport paths may be used only with checked target, status, and approval.
False assumption: a reachable external tool is automatically part of the allowed Nova tool chain.
Target, transport status, import or connection approval, and expected data movement must be visible.
Do not silently add third-party tools; check current approval, target, and response status before use.
Recheck if the gateway, proxy, remote source, or target system changes or emits warnings.
Safety Axes
How this path can affect work
Axes are stable catalog signals for humans, agents, and LLM discovery. One path can carry several axes.
read_current_state
Reads current state, response signals, or evidence without treating that alone as permission for a follow-up action.
Use the signal as current evidence and re-check target, scope, and visible state before any follow-up action.external_connection
Uses proxy, external MCP gateways, network routes, third-party integrations, or import paths.
Check external target, transport status, and approval before use; do not silently chain third-party tools.automation_run
Starts or monitors crawls, sequences, schedulers, tasks, batches, or longer runs.
Keep scope, budget, progress, stop condition, and terminal status visible before and during the run.user_confirmation
Requires visible confirmation, target review, approval, or deliberate user control.
Do not proceed until the required confirmation is visible or unambiguous in the current context.