Sliq

Start Find Search

start_find_search

Kick off an Exa Websets search for a named list on an agent.

An agent can host many named lists. Each list spawns its own Exa webset, so distinct criteria buckets (e.g. 'oil & gas operators' vs 'midstream') don't share Exa's per-webset domain-dedup.

Costs 2 Sliq credits per ingested result (charged during ingest). Free when the user has their own Exa key connected.

Refuses to start when the balance can't afford even one result in sliq_credits mode. When the balance covers fewer results than requested, the count is clamped to what's affordable and the result carries a clamp_notice — surface it to the user ("you have X credits, so I'll find Y").

Input Schema

TableJSON Schema

Name	Required	Description
`agent_id`	Yes	The agent to host the new list.
`criteria`	Yes	what to search for (the Exa query body).
`excluded`	No	company domains/URLs the webset must not ingest (compared by canonical hostname). Use to avoid re-ingesting rows already recorded in the list, e.g. seeds from find_companies_by_tech_stack.
`list_name`	Yes	short kebab slug identifying the bucket (e.g. 'oil-gas-operators'). Pick a name distinct from any existing list on the agent. Re-using a finished list's name extends it; re-using a running list's name is refused.
`entity_type`	Yes	'person' or 'company'. Mixed people+company agents are supported — each list has its own row type.
`max_results`	No	target count (capped at the per-list provider cap).
`people_query`	No	For company lists, a positive-form query for the people to attach as a `data.people[]` block on each company row — controls which people come back. Company lists only; skipped on person lists.
`people_titles`	No	The roles `people_query` is asking for, as title keywords, e.g. ['Marketing', 'Growth'] or ['Operations', 'Chief']. Each is matched as a case-insensitive substring of a person's job title and the list is OR-ed, so use the distinctive word a title is built from rather than a whole title: companies word the same job many ways, and a full phrase finds only the ones that word it exactly as written here. Broader words recall more people and admit more adjacent roles, so pick the narrowest word that still covers the variants. Set it whenever `people_query` names roles — the attachment runs unattended, so this is its only structured read of the roles you asked for. Company lists only.
`people_preferences`	No	Soft ranking/filtering preferences (and exclusions) for the attached people. Company lists only.
`people_exclude_titles`	No	Title keywords to reject, e.g. ['Coordinator', 'Intern']. Same substring matching as `people_titles`, subtracted from it. Whenever `people_preferences` rules a role out, put that role here too — the attachment cannot read the prose, so an exclusion written only there is not applied. Company lists only.
`people_include_alumni`	No	Include past employees (alumni) alongside current ones in the attached people block (default false). Company lists only.

Schema Changelog

Changes observed during successful MCP inspections.

Changed5 schema fields changedSep 23, 2026

addedInput schema / properties / agent_id

Added value: +{
+  "description": "The agent to host the new list.",
+  "type": "integer"
+}

changedInput schema / properties / entity_type / description

Previous value: -"'person' or 'company'. Mixed people+company tasks are\nsupported — each list has its own row type."New value: +"'person' or 'company'. Mixed people+company agents are\nsupported — each list has its own row type."

changedInput schema / properties / list_name / description

Previous value: -"short kebab slug identifying the bucket (e.g.\n'oil-gas-operators'). Pick a name distinct from any existing list\non the task. Re-using a finished list's name extends it; re-using\na running list's name is refused."New value: +"short kebab slug identifying the bucket (e.g.\n'oil-gas-operators'). Pick a name distinct from any existing list\non the agent. Re-using a finished list's name extends it; re-using\na running list's name is refused."

removedInput schema / properties / task_id

Removed value: -{
-  "description": "The task to host the new list.",
-  "type": "integer"
-}

changedInput schema / required

Previous value: -[
-  "task_id",
-  "list_name",
-  "criteria",
-  "entity_type"
-]New value: +[
+  "agent_id",
+  "list_name",
+  "criteria",
+  "entity_type"
+]

First observedSep 15, 2026

TDQS

A4/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With only readOnly=false and destructive=false in annotations, the description carries the economic and failure-mode burden and does it well: charges 2 Sliq credits per ingested result, free with a user Exa key, refuses when unaffordable, and clamps the count with a clamp_notice. It also explains per-webset domain-dedup behavior. No contradiction with annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four short paragraphs, each earning its place: purpose, dedup rationale, cost model, and failure/clamping behavior. It is front-loaded with a single-sentence purpose and does not repeat schema content.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a complex async kick-off tool, it covers economics, failure modes, clamping, and exactly what to surface to the user. It stops short of describing the response shape or how to retrieve/monitor results, which is more notable because no output schema exists.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the schema descriptions are already very detailed, so the baseline is 3. The description adds context about list buckets and credit clamping but does not add syntax or meaning for individual parameters beyond what the schema already provides.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific action ('Kick off an Exa Websets search') and a specific resource ('a named list on an agent'), so the core behavior is unmistakable. It doesn't explicitly name siblings like extend_find_search or record_search_results, so differentiation comes from the WebSets/list terminology rather than an explicit contrast.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives useful context for when to create separate lists (per-webset domain-dedup isolation) and notes list-name reuse constraints, which implies starting vs extending behavior. It never names an alternative tool or states when not to use this one, leaving the vs-alternatives question to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Sliq tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?