Struct WebSearchExecutor 

pub struct WebSearchExecutor { /* private fields */ }

Issues a natural-language query to an external search API and returns ranked results.

Security

• The search endpoint is validated with the same SSRF machinery as WebScrapeExecutor: HTTPS-only, DNS-resolved and checked against private ranges, and the resolved addresses are pinned into the reqwest::Client via resolve_to_addrs to close the DNS-rebinding TOCTOU window.
• [tools.scrape].denied_domains is enforced against the endpoint; the scrape allowlist is intentionally not consulted (the endpoint is operator-configured infrastructure, not an LLM-chosen target).
• Rendered result text (titles + snippets) passes through the IPI filter before reaching the LLM, since snippet content originates from arbitrary indexed pages.
• Result URLs are returned as text only — never auto-fetched by this tool.

Example

use zeph_tools::{SearchConfig, ScrapeConfig};
use zeph_tools::search::WebSearchExecutor;
use zeph_common::secret::Secret;

let cfg = SearchConfig { enabled: true, ..SearchConfig::default() };
let executor = WebSearchExecutor::new(&cfg, &ScrapeConfig::default(), Some(Secret::new("key")));
assert!(executor.is_some());

Implementations

impl WebSearchExecutor

pub fn new( cfg: &SearchConfig, scrape_cfg: &ScrapeConfig, api_key: Option<Secret>, ) -> Option<Self>

Build a WebSearchExecutor from configuration.

Returns Some only when cfg.enabled is true AND SearchBackend::from_config succeeds (e.g. a valid key is present for a keyed backend). Returns None otherwise — the caller must omit the tool from the executor chain entirely in that case, so tool_definitions() never advertises an unusable tool to the LLM (FR-002).

denied_domains and ipi_filter_threshold are read from [tools.scrape] — the search tool has no independent domain-policy or IPI-threshold configuration.

No network connections are made at construction time.

pub fn with_audit(self, logger: Arc<AuditLogger>) -> Self

Attach an audit logger. Each tool invocation will emit an AuditEntry.

pub fn with_egress_config(self, config: EgressConfig) -> Self

Configure egress event logging.

pub fn with_egress_tx( self, tx: Sender<EgressEvent>, dropped: Arc<AtomicU64>, ) -> Self

Attach the egress telemetry channel sender and drop counter.

pub fn egress_dropped(&self) -> Arc<AtomicU64>

Returns a clone of the egress drop counter, for use in the drain task.

Trait Implementations

impl Debug for WebSearchExecutor

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl ToolExecutor for WebSearchExecutor

fn tool_definitions(&self) -> Vec<ToolDef>

Return the tool definitions this executor can handle.

async fn execute( &self, _response: &str, ) -> Result<Option<ToolOutput>, ToolError>

Parse response for fenced tool blocks and execute them.

async fn execute_tool_call( &self, call: &ToolCall, ) -> Result<Option<ToolOutput>, ToolError>

Execute a structured tool call. Returns Ok(None) if call.tool_id is not handled.

fn is_tool_retryable(&self, tool_id: &str) -> bool

Whether the executor can safely retry this tool call on a transient error.

fn requires_confirmation(&self, _call: &ToolCall) -> bool

Return true when call would require user confirmation before execution.

fn execute_tool_call_confirmed( &self, call: &ToolCall, ) -> impl Future<Output = Result<Option<ToolOutput>, ToolError>> + Send

Execute a structured tool call bypassing confirmation checks.

fn checkpoint_undo(&self, _n: usize) -> CheckpointActionResult

Undo the last n checkpointed write commands.

fn checkpoint_redo(&self) -> CheckpointActionResult

Redo the last undone checkpoint.

fn checkpoint_list(&self) -> CheckpointListResult

List the current undo stack entries and redo depth.

fn is_tool_speculatable(&self, _tool_id: &str) -> bool

Whether a tool call can be safely dispatched speculatively (before the LLM finishes).

fn execute_confirmed( &self, response: &str, ) -> impl Future<Output = Result<Option<ToolOutput>, ToolError>> + Send

Execute bypassing confirmation checks (called after user approves).

fn set_skill_env(&self, _env: Option<HashMap<String, String>>)

Inject environment variables for the currently active skill. No-op by default.

fn set_effective_trust(&self, _level: SkillTrustLevel)

Set the effective trust level for the currently active skill. No-op by default.