exa-policy-guardrails
Implement content policy enforcement, domain filtering, and usage guardrails for Exa. Use when setting up content safety rules, restricting search domains, or enforcing query and budget policies for Exa integrations. Trigger with phrases like "exa policy", "exa content filter", "exa guardrails", "exa domain allowlist", "exa content moderation".
What this skill does
# Exa Policy Guardrails
## Overview
Policy enforcement for Exa neural search integrations. Exa searches the open web, so results may include unreliable sources, competitor content, or inappropriate material. This skill covers domain allowlists/blocklists (via Exa's `includeDomains`/`excludeDomains`), content moderation, query sanitization, freshness policies, and per-user budget enforcement.
## Prerequisites
- `exa-js` installed and configured
- Content policy requirements defined
- Redis for per-user quota tracking (optional)
## Instructions
### Step 1: Domain Filtering (Built-in Exa Feature)
```typescript
import Exa from "exa-js";
const exa = new Exa(process.env.EXA_API_KEY);
// Exa supports up to 1200 domains in includeDomains/excludeDomains
const TRUSTED_SOURCES = {
medical: [
"pubmed.ncbi.nlm.nih.gov", "who.int", "cdc.gov",
"nejm.org", "nature.com", "thelancet.com",
],
technical: [
"github.com", "stackoverflow.com", "developer.mozilla.org",
"docs.python.org", "nodejs.org", "arxiv.org",
],
news: [
"reuters.com", "apnews.com", "bbc.com",
"techcrunch.com", "arstechnica.com",
],
};
const BLOCKED_DOMAINS = [
"competitor1.com", "competitor2.io",
"spam-farm.com", "content-mill.net",
];
async function policySearch(
query: string,
category: keyof typeof TRUSTED_SOURCES | "general"
) {
const opts: any = {
type: "auto",
numResults: 10,
text: { maxCharacters: 1000 },
moderation: true, // Exa's built-in content moderation
};
if (category !== "general" && TRUSTED_SOURCES[category]) {
opts.includeDomains = TRUSTED_SOURCES[category];
} else {
opts.excludeDomains = BLOCKED_DOMAINS;
}
return exa.searchAndContents(query, opts);
}
```
### Step 2: Query Content Policy
```typescript
const BLOCKED_PATTERNS = [
/how to (hack|exploit|attack|ddos)/i,
/(buy|purchase|order)\s+(drugs|weapons|firearms)/i,
/personal.*(address|phone|ssn|social security)/i,
/generate.*(malware|ransomware|virus)/i,
];
function validateQuery(input: string): string {
for (const pattern of BLOCKED_PATTERNS) {
if (pattern.test(input)) {
throw new PolicyViolation("Query blocked by content policy");
}
}
// Sanitize
return input
.replace(/[<>{}]/g, "") // strip HTML/template chars
.replace(/\0/g, "") // remove null bytes
.trim()
.substring(0, 500); // cap query length
}
class PolicyViolation extends Error {
constructor(message: string) {
super(message);
this.name = "PolicyViolation";
}
}
```
### Step 3: Freshness Policy
```typescript
// Enforce minimum recency for time-sensitive use cases
function applyFreshnessPolicy(
opts: any,
maxAgeDays: number
): any {
const cutoff = new Date(Date.now() - maxAgeDays * 24 * 60 * 60 * 1000);
return {
...opts,
startPublishedDate: cutoff.toISOString(),
};
}
// Usage: only return results from the last 90 days
const results = await exa.searchAndContents("AI regulation updates",
applyFreshnessPolicy(
{ type: "neural", numResults: 10, text: true },
90 // max 90 days old
)
);
```
### Step 4: Per-User Budget Enforcement
```typescript
class ExaUsagePolicy {
private usage = new Map<string, { count: number; resetAt: number }>();
private limits: Record<string, number>;
constructor(limits: Record<string, number> = {
"free": 10,
"pro": 100,
"enterprise": 1000,
}) {
this.limits = limits;
}
checkQuota(userId: string, tier: string): void {
const limit = this.limits[tier] || this.limits["free"] || 10;
const now = Date.now();
const hourKey = `${userId}:${new Date().toISOString().substring(0, 13)}`;
let entry = this.usage.get(hourKey);
if (!entry || entry.resetAt < now) {
entry = { count: 0, resetAt: now + 3600 * 1000 };
}
if (entry.count >= limit) {
throw new PolicyViolation(
`Hourly search quota exceeded: ${entry.count}/${limit}`
);
}
entry.count++;
this.usage.set(hourKey, entry);
}
}
const usagePolicy = new ExaUsagePolicy();
```
### Step 5: Combined Policy Enforcement
```typescript
async function enforcedSearch(
userId: string,
userTier: string,
rawQuery: string,
category: keyof typeof TRUSTED_SOURCES | "general" = "general",
maxAgeDays?: number
) {
// 1. Check quota
usagePolicy.checkQuota(userId, userTier);
// 2. Validate and sanitize query
const query = validateQuery(rawQuery);
// 3. Build options with domain policy
let opts: any = {
type: "auto",
numResults: 10,
text: { maxCharacters: 1000 },
moderation: true,
};
if (category !== "general" && TRUSTED_SOURCES[category]) {
opts.includeDomains = TRUSTED_SOURCES[category];
} else {
opts.excludeDomains = BLOCKED_DOMAINS;
}
// 4. Apply freshness policy
if (maxAgeDays) {
opts = applyFreshnessPolicy(opts, maxAgeDays);
}
// 5. Execute search
return exa.searchAndContents(query, opts);
}
```
## Error Handling
| Issue | Cause | Solution |
|-------|-------|----------|
| Competitor content in results | No domain filtering | Apply `excludeDomains` blocklist |
| Harmful query accepted | No content policy | Validate queries against blocked patterns |
| Stale results displayed | No freshness check | Apply `startPublishedDate` filter |
| API cost overrun | No usage limits | Implement per-user/tier quotas |
| Blocked policy query | False positive | Review and adjust `BLOCKED_PATTERNS` |
## Resources
- [Exa Search Reference](https://docs.exa.ai/reference/search)
- [Exa Domain Filtering](https://docs.exa.ai/reference/search)
## Next Steps
For architecture decisions, see `exa-architecture-variants`. For cost control, see `exa-cost-tuning`.
Related in Writing & Docs
jax-development
IncludedUse this skill when the user is writing, debugging, profiling, refactoring, reviewing, benchmarking, parallelising, exporting, or explaining JAX code, or when they mention JAX, jax.numpy, jit, grad, value_and_grad, vmap, scan, lax, random keys, pytrees, jax.Array, sharding, Mesh, PartitionSpec, NamedSharding, pmap, shard_map, Pallas, XLA, StableHLO, checkify, profiler, or the JAX repo. It helps turn NumPy or PyTorch-style code into pure functional JAX, fix tracer/control-flow/shape/PRNG bugs, remove recompiles and host-device syncs, choose transforms and sharding strategies, inspect jaxpr/lowering/IR, and benchmark compiled code correctly.
nature-article-writer
IncludedDrafts, rewrites, diagnostically critiques, and style-calibrates primary research manuscripts for Nature and Nature Portfolio journals. Use when the user wants a Nature-style title, summary paragraph or abstract, introduction, results, discussion, methods, figure legends, presubmission enquiry, cover letter, reviewer response, or when a scientific draft sounds generic, jargon-heavy, structurally weak, or AI-ish and needs precise, broad-reader-friendly prose without inventing data, analyses, or references. Best for primary research articles and letters rather than reviews or press releases unless explicitly adapting one.
deckrd
IncludedDocument-driven framework that derives requirements, specifications, implementation plans, and executable tasks from goals through structured AI dialogue. Use when user says "write requirements", "create spec", "plan implementation", "derive tasks", "structure this feature", "break down into tasks", or "document this module". Also use for reverse engineering existing code into docs (/deckrd rev). Do NOT use for direct code writing — use /deckrd-coder after tasks are generated. Do NOT use when the user only wants to run or fix existing code without planning.
clinical-decision-support
IncludedGenerate professional clinical decision support (CDS) documents for pharmaceutical and clinical research settings, including patient cohort analyses (biomarker-stratified with outcomes) and treatment recommendation reports (evidence-based guidelines with decision algorithms). Supports GRADE evidence grading, statistical analysis (hazard ratios, survival curves, waterfall plots), biomarker integration, and regulatory compliance. Outputs publication-ready LaTeX/PDF format optimized for drug development, clinical research, and evidence synthesis.
handling-sf-data
IncludedSalesforce data operations with 130-point scoring. Use this skill to create, update, delete, bulk import/export, generate test data, and clean up org records using sf CLI and anonymous Apex. TRIGGER when: user creates test data, performs bulk import/export, uses sf data CLI commands, needs data factory patterns for Apex tests, or needs to seed/clean records in a Salesforce org. DO NOT TRIGGER when: SOQL query writing only (use querying-soql), Apex test execution (use running-apex-tests), or metadata deployment (use deploying-metadata).
accelint-ac-to-playwright
IncludedConvert and validate acceptance criteria for Playwright test automation. Use when user asks to (1) review/evaluate/check if AC are ready for automation, (2) assess if AC can be converted as-is, (3) validate AC quality for Playwright, (4) turn AC into tests, (5) generate tests from acceptance criteria, (6) convert .md bullets or .feature Gherkin files to Playwright specs, (7) create test automation from requirements. Handles both bullet-style markdown and Gherkin syntax with JSON test plan generation and validation.