Evaluate caller-provided candidate context and return decision-ready output. This is the primary FreshContext judgment path: it does not fetch, crawl, scrape, browse, read folders, or call adapters.
AI agents call evaluate_context to retrieve information from Freshcontext without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
now | string | — | Optional ISO timestamp for deterministic evaluation. |
intent | string | Yes | Intent Profile id, e.g. citation_check, student_research, developer_adoption, job_search, market_watch, business_due_diligence, medical_literature_triage. |
profile | string | Yes | Source Profile id, e.g. academic_research, jobs_opportunities, market_finance, official_docs, local_custom. |
signals | array | Yes | Candidate context items provided by the caller. FreshContext evaluates these; it does not retrieve them. |
Parameters from the server's own tool schema.
This is a pure evaluation/analysis function that takes input provided by the caller and produces analysis output. It has no side effects, does not access external systems, and does not modify any state.
From the tool's definition Tool description explicitly states it 'does not fetch, crawl, scrape, browse, read folders, or call adapters' and only 'Evaluate[s] caller-provided candidate context and return[s] decision-ready output.' The tool processes data already supplied by the caller…
Risk signalsAccepts raw HTML/template content (signals[].content) · High parameter count (17 properties)
Documented attack patterns abuse exactly the kind of access evaluate_context gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Freshcontext, and nothing reaches the server without passing your rules. This is the rule we recommend for evaluate_context:
{
"version": "1",
"default": "deny",
"tools": {
"evaluate_context": {}
}
}evaluate_context is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Evaluate caller-provided candidate context and return decision-ready output. This is the primary FreshContext judgment path: it does not fetch, crawl, scrape, browse, read folders, or call adapters. It is categorised as a Read tool in the Freshcontext MCP Server, which means it retrieves data without modifying state.
evaluate_context accepts 4 parameters: now, intent, profile, signals. Required: intent, profile, signals. The full parameter table on this page comes from the server's own tool schema.
Register the Freshcontext MCP server in PolicyLayer and add a rule for evaluate_context: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Freshcontext. Nothing to install.
evaluate_context is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the evaluate_context rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for evaluate_context. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
evaluate_context is provided by the Freshcontext MCP server (PrinceGabriel-lgtm/freshcontext-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Freshcontext, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
22 Freshcontext tools catalogued and risk-classified — across an index of 43,000+ MCP servers.