LinkedIn Ads MCP Server

run_full_audit

Identify and fix common LinkedIn Ads campaign issues with a 10-point audit covering declining CTR, low budget, and missing tracking.

Instructions

Run a comprehensive audit across 10 checks: declining CTR, low budget utilization, missing conversion tracking, maximized delivery, audience expansion, audience network, stale campaigns, creative fatigue, overlapping audiences, missing UTM parameters.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`lookbackDays`	No	Days of data to analyze (default 28)

Implementation Reference

src/tools/audit.ts:51-305 (handler)

The handler function for the run_full_audit tool. It fetches all active campaigns, analytics, and conversion rules, then runs 10 audit checks (declining CTR, low budget utilization, missing conversion tracking, maximized delivery, audience expansion, audience network, stale campaigns, creative fatigue, overlapping audiences, missing UTM parameters) and returns an AuditReport with all findings.

handler: async (args: { lookbackDays: number }) => {
  const findings: AuditFinding[] = [];

  // Pull all active campaigns
  const campaignsPath = client.accountPath("/adCampaigns");
  const campaignsRes = await client.get<{
    elements: Array<{
      id: string;
      name: string;
      status: string;
      enableAudienceExpansion?: boolean;
      offsiteDeliveryEnabled?: boolean;
      bidStrategy?: string;
      dailyBudget?: { amount: string };
      targetingCriteria?: Record<string, unknown>;
    }>;
  }>(campaignsPath, {
    q: "search",
    "search.status.values[0]": "ACTIVE",
  });
  const campaigns = campaignsRes.data.elements || [];

  // Pull analytics for the lookback period
  const endDate = new Date();
  const startDate = new Date();
  startDate.setDate(endDate.getDate() - args.lookbackDays);
  const fmt = (d: Date) => d.toISOString().split("T")[0];

  const [startYear, startMonth, startDay] = fmt(startDate).split("-");
  const [endYear, endMonth, endDay] = fmt(endDate).split("-");

  let analyticsElements: Array<Record<string, unknown>> = [];
  if (campaigns.length > 0) {
    const analyticsParams: Record<string, string> = {
      q: "analytics",
      pivot: "CAMPAIGN",
      timeGranularity: "DAILY",
      "dateRange.start.year": startYear,
      "dateRange.start.month": startMonth,
      "dateRange.start.day": startDay,
      "dateRange.end.year": endYear,
      "dateRange.end.month": endMonth,
      "dateRange.end.day": endDay,
      fields: "impressions,clicks,costInLocalCurrency,conversions",
    };
    campaigns.slice(0, 20).forEach((c, i) => {
      analyticsParams[`campaigns[${i}]`] = `urn:li:sponsoredCampaign:${c.id}`;
    });

    const analyticsRes = await client.get<{
      elements: Array<Record<string, unknown>>;
    }>("/adAnalytics", analyticsParams);
    analyticsElements = analyticsRes.data.elements || [];
  }

  // Pull conversion rules
  const conversionsRes = await client.get<{
    elements: Array<{ id: string; campaigns?: string[] }>;
  }>("/conversions", {
    q: "account",
    account: client.config.adAccountId,
  });
  const conversionRules = conversionsRes.data.elements || [];

  // ── Check 1: Declining CTR ──────────────────────────────
  const weeklyCtrs = computeWeeklyCtr(analyticsElements);
  if (weeklyCtrs.length >= 3) {
    const declining = isConsistentlyDeclining(weeklyCtrs);
    if (declining) {
      findings.push({
        check: "Declining CTR",
        severity: "warning",
        affectedEntities: ["account-wide"],
        description: `CTR has declined for ${weeklyCtrs.length} consecutive weeks: ${weeklyCtrs.map((c) => (c * 100).toFixed(2) + "%").join(" → ")}`,
        recommendation:
          "Refresh ad creatives, test new messaging, or narrow targeting to more relevant audiences.",
      });
    }
  }

  // ── Check 2: Low budget utilization ─────────────────────
  for (const campaign of campaigns) {
    if (!campaign.dailyBudget) continue;
    const dailyBudget = parseFloat(campaign.dailyBudget.amount);
    const campaignAnalytics = analyticsElements.filter(
      (a) => String(a.pivotValue).includes(campaign.id)
    );
    const totalSpend = campaignAnalytics.reduce(
      (sum, a) => sum + (Number(a.costInLocalCurrency) || 0),
      0
    );
    const avgDailySpend = totalSpend / Math.max(args.lookbackDays, 1);
    const utilization = avgDailySpend / dailyBudget;

    if (utilization < 0.7) {
      findings.push({
        check: "Low Budget Utilization",
        severity: "warning",
        affectedEntities: [campaign.id],
        description: `Campaign "${campaign.name}" is only utilizing ${(utilization * 100).toFixed(0)}% of its $${dailyBudget}/day budget.`,
        recommendation:
          "Broaden targeting, increase bids, or reduce the daily budget to match actual delivery.",
      });
    }
  }

  // ── Check 3: Missing conversion tracking ────────────────
  const campaignsWithConversions = new Set(
    conversionRules.flatMap((r) => r.campaigns || [])
  );
  const missingConversions = campaigns.filter(
    (c) => !campaignsWithConversions.has(`urn:li:sponsoredCampaign:${c.id}`)
  );
  if (missingConversions.length > 0) {
    findings.push({
      check: "Missing Conversion Tracking",
      severity: "critical",
      affectedEntities: missingConversions.map((c) => c.id),
      description: `${missingConversions.length} active campaign(s) have no conversion tracking configured.`,
      recommendation:
        "Add LinkedIn Insight Tag conversion events to track ROI. Without conversion tracking, optimization is impossible.",
    });
  }

  // ── Check 4: Maximized delivery (automated bidding) ─────
  const autoBidCampaigns = campaigns.filter(
    (c) => c.bidStrategy === "MAXIMUM_DELIVERY"
  );
  if (autoBidCampaigns.length > 0) {
    findings.push({
      check: "Maximized Delivery (Automated Bidding)",
      severity: "warning",
      affectedEntities: autoBidCampaigns.map((c) => c.id),
      description: `${autoBidCampaigns.length} campaign(s) use MAXIMUM_DELIVERY bidding, giving LinkedIn full control over bid prices.`,
      recommendation:
        "Switch to MANUAL bidding to control CPCs/CPMs. Automated bidding often overspends, especially in competitive auctions.",
    });
  }

  // ── Check 5: Audience expansion enabled ─────────────────
  const expansionOn = campaigns.filter((c) => c.enableAudienceExpansion);
  if (expansionOn.length > 0) {
    findings.push({
      check: "Audience Expansion Enabled",
      severity: "critical",
      affectedEntities: expansionOn.map((c) => c.id),
      description: `${expansionOn.length} campaign(s) have audience expansion ON — LinkedIn will show ads to people outside your targeting criteria.`,
      recommendation:
        "Disable audience expansion immediately. It dilutes targeting precision and wastes budget on irrelevant impressions.",
    });
  }

  // ── Check 6: Audience network enabled ───────────────────
  const networkOn = campaigns.filter((c) => c.offsiteDeliveryEnabled);
  if (networkOn.length > 0) {
    findings.push({
      check: "Audience Network Enabled",
      severity: "critical",
      affectedEntities: networkOn.map((c) => c.id),
      description: `${networkOn.length} campaign(s) have LinkedIn Audience Network ON — ads are being served on third-party sites.`,
      recommendation:
        "Disable the Audience Network. Offsite placements have significantly lower engagement and conversion rates.",
    });
  }

  // ── Check 7: Stale campaigns ────────────────────────────
  const recentDays = 7;
  const recentStart = new Date();
  recentStart.setDate(recentStart.getDate() - recentDays);
  const staleCampaigns = campaigns.filter((c) => {
    const campaignData = analyticsElements.filter((a) =>
      String(a.pivotValue).includes(c.id)
    );
    const recentImpressions = campaignData
      .filter((a) => {
        const dateStr = a.dateRange as { start?: { year: number; month: number; day: number } } | undefined;
        if (!dateStr?.start) return false;
        const d = new Date(dateStr.start.year, dateStr.start.month - 1, dateStr.start.day);
        return d >= recentStart;
      })
      .reduce((sum, a) => sum + (Number(a.impressions) || 0), 0);
    return recentImpressions === 0;
  });
  if (staleCampaigns.length > 0) {
    findings.push({
      check: "Stale Campaigns",
      severity: "info",
      affectedEntities: staleCampaigns.map((c) => c.id),
      description: `${staleCampaigns.length} active campaign(s) have delivered zero impressions in the last ${recentDays} days.`,
      recommendation:
        "Pause or archive campaigns with no delivery. Check if targeting is too narrow, bids are too low, or budgets are exhausted.",
    });
  }

  // ── Check 8: Creative fatigue ───────────────────────────
  // Simplified: flag if overall CTR in the last week is < 70% of the first week
  if (weeklyCtrs.length >= 2) {
    const firstWeekCtr = weeklyCtrs[0];
    const lastWeekCtr = weeklyCtrs[weeklyCtrs.length - 1];
    if (firstWeekCtr > 0 && lastWeekCtr / firstWeekCtr < 0.7) {
      findings.push({
        check: "Creative Fatigue",
        severity: "warning",
        affectedEntities: ["account-wide"],
        description: `CTR has dropped ${((1 - lastWeekCtr / firstWeekCtr) * 100).toFixed(0)}% from the first week (${(firstWeekCtr * 100).toFixed(2)}% → ${(lastWeekCtr * 100).toFixed(2)}%).`,
        recommendation:
          "Rotate in fresh creatives. Audiences seeing the same ads repeatedly leads to banner blindness and declining engagement.",
      });
    }
  }

  // ── Check 9: Overlapping audiences ──────────────────────
  const targetingGroups = campaigns
    .filter((c) => c.targetingCriteria)
    .map((c) => ({
      id: c.id,
      targeting: JSON.stringify(c.targetingCriteria),
    }));
  const duplicates = findDuplicateTargeting(targetingGroups);
  if (duplicates.length > 0) {
    findings.push({
      check: "Overlapping Audiences",
      severity: "warning",
      affectedEntities: duplicates.flat(),
      description: `${duplicates.length} pair(s) of campaigns share identical targeting criteria, causing self-competition in the auction.`,
      recommendation:
        "Differentiate targeting between campaigns or consolidate them. Overlapping audiences drive up your own CPCs.",
    });
  }

  // ── Check 10: Missing UTM parameters ────────────────────
  // This would require pulling creatives — simplified check
  findings.push({
    check: "UTM Parameters",
    severity: "info",
    affectedEntities: ["requires-creative-scan"],
    description:
      "UTM parameter check requires scanning creative destination URLs. Use list_creatives to audit individually.",
    recommendation:
      "Ensure all destination URLs include utm_source=linkedin&utm_medium=paid_social&utm_campaign={campaign_name}.",
  });

  // ── Compile report ──────────────────────────────────────
  const report: AuditReport = {
    accountId: client.config.adAccountId,
    timestamp: new Date().toISOString(),
    totalFindings: findings.length,
    critical: findings.filter((f) => f.severity === "critical").length,
    warnings: findings.filter((f) => f.severity === "warning").length,
    info: findings.filter((f) => f.severity === "info").length,
    findings,
  };

  return report;
},

src/tools/audit.ts:45-49 (schema)
Input schema for run_full_audit: accepts a single optional parameter 'lookbackDays' (number, default 28) describing how many days of data to analyze.
```
schema: z.object({
  lookbackDays: z
    .number()
    .default(28)
    .describe("Days of data to analyze (default 28)"),
```
src/index.ts:108-108 (registration)
Registration of audit tools (including run_full_audit) via registerAuditTools(linkedIn) call, returning an object keyed by tool name that gets spread into the MCP server's tool groups.
```
...registerAuditTools(linkedIn),
```

src/tools/audit.ts:38-335 (registration)

The registerAuditTools function that defines the tool object with description, schema, and handler. The returned object is spread into the MCP server's tool registry.

export function registerAuditTools(client: LinkedInClient) {
  return {
    run_full_audit: {
      description:
        "Run a comprehensive audit across 10 checks: declining CTR, low budget utilization, " +
        "missing conversion tracking, maximized delivery, audience expansion, audience network, " +
        "stale campaigns, creative fatigue, overlapping audiences, missing UTM parameters.",
      schema: z.object({
        lookbackDays: z
          .number()
          .default(28)
          .describe("Days of data to analyze (default 28)"),
      }),
      handler: async (args: { lookbackDays: number }) => {
        const findings: AuditFinding[] = [];

        // Pull all active campaigns
        const campaignsPath = client.accountPath("/adCampaigns");
        const campaignsRes = await client.get<{
          elements: Array<{
            id: string;
            name: string;
            status: string;
            enableAudienceExpansion?: boolean;
            offsiteDeliveryEnabled?: boolean;
            bidStrategy?: string;
            dailyBudget?: { amount: string };
            targetingCriteria?: Record<string, unknown>;
          }>;
        }>(campaignsPath, {
          q: "search",
          "search.status.values[0]": "ACTIVE",
        });
        const campaigns = campaignsRes.data.elements || [];

        // Pull analytics for the lookback period
        const endDate = new Date();
        const startDate = new Date();
        startDate.setDate(endDate.getDate() - args.lookbackDays);
        const fmt = (d: Date) => d.toISOString().split("T")[0];

        const [startYear, startMonth, startDay] = fmt(startDate).split("-");
        const [endYear, endMonth, endDay] = fmt(endDate).split("-");

        let analyticsElements: Array<Record<string, unknown>> = [];
        if (campaigns.length > 0) {
          const analyticsParams: Record<string, string> = {
            q: "analytics",
            pivot: "CAMPAIGN",
            timeGranularity: "DAILY",
            "dateRange.start.year": startYear,
            "dateRange.start.month": startMonth,
            "dateRange.start.day": startDay,
            "dateRange.end.year": endYear,
            "dateRange.end.month": endMonth,
            "dateRange.end.day": endDay,
            fields: "impressions,clicks,costInLocalCurrency,conversions",
          };
          campaigns.slice(0, 20).forEach((c, i) => {
            analyticsParams[`campaigns[${i}]`] = `urn:li:sponsoredCampaign:${c.id}`;
          });

          const analyticsRes = await client.get<{
            elements: Array<Record<string, unknown>>;
          }>("/adAnalytics", analyticsParams);
          analyticsElements = analyticsRes.data.elements || [];
        }

        // Pull conversion rules
        const conversionsRes = await client.get<{
          elements: Array<{ id: string; campaigns?: string[] }>;
        }>("/conversions", {
          q: "account",
          account: client.config.adAccountId,
        });
        const conversionRules = conversionsRes.data.elements || [];

        // ── Check 1: Declining CTR ──────────────────────────────
        const weeklyCtrs = computeWeeklyCtr(analyticsElements);
        if (weeklyCtrs.length >= 3) {
          const declining = isConsistentlyDeclining(weeklyCtrs);
          if (declining) {
            findings.push({
              check: "Declining CTR",
              severity: "warning",
              affectedEntities: ["account-wide"],
              description: `CTR has declined for ${weeklyCtrs.length} consecutive weeks: ${weeklyCtrs.map((c) => (c * 100).toFixed(2) + "%").join(" → ")}`,
              recommendation:
                "Refresh ad creatives, test new messaging, or narrow targeting to more relevant audiences.",
            });
          }
        }

        // ── Check 2: Low budget utilization ─────────────────────
        for (const campaign of campaigns) {
          if (!campaign.dailyBudget) continue;
          const dailyBudget = parseFloat(campaign.dailyBudget.amount);
          const campaignAnalytics = analyticsElements.filter(
            (a) => String(a.pivotValue).includes(campaign.id)
          );
          const totalSpend = campaignAnalytics.reduce(
            (sum, a) => sum + (Number(a.costInLocalCurrency) || 0),
            0
          );
          const avgDailySpend = totalSpend / Math.max(args.lookbackDays, 1);
          const utilization = avgDailySpend / dailyBudget;

          if (utilization < 0.7) {
            findings.push({
              check: "Low Budget Utilization",
              severity: "warning",
              affectedEntities: [campaign.id],
              description: `Campaign "${campaign.name}" is only utilizing ${(utilization * 100).toFixed(0)}% of its $${dailyBudget}/day budget.`,
              recommendation:
                "Broaden targeting, increase bids, or reduce the daily budget to match actual delivery.",
            });
          }
        }

        // ── Check 3: Missing conversion tracking ────────────────
        const campaignsWithConversions = new Set(
          conversionRules.flatMap((r) => r.campaigns || [])
        );
        const missingConversions = campaigns.filter(
          (c) => !campaignsWithConversions.has(`urn:li:sponsoredCampaign:${c.id}`)
        );
        if (missingConversions.length > 0) {
          findings.push({
            check: "Missing Conversion Tracking",
            severity: "critical",
            affectedEntities: missingConversions.map((c) => c.id),
            description: `${missingConversions.length} active campaign(s) have no conversion tracking configured.`,
            recommendation:
              "Add LinkedIn Insight Tag conversion events to track ROI. Without conversion tracking, optimization is impossible.",
          });
        }

        // ── Check 4: Maximized delivery (automated bidding) ─────
        const autoBidCampaigns = campaigns.filter(
          (c) => c.bidStrategy === "MAXIMUM_DELIVERY"
        );
        if (autoBidCampaigns.length > 0) {
          findings.push({
            check: "Maximized Delivery (Automated Bidding)",
            severity: "warning",
            affectedEntities: autoBidCampaigns.map((c) => c.id),
            description: `${autoBidCampaigns.length} campaign(s) use MAXIMUM_DELIVERY bidding, giving LinkedIn full control over bid prices.`,
            recommendation:
              "Switch to MANUAL bidding to control CPCs/CPMs. Automated bidding often overspends, especially in competitive auctions.",
          });
        }

        // ── Check 5: Audience expansion enabled ─────────────────
        const expansionOn = campaigns.filter((c) => c.enableAudienceExpansion);
        if (expansionOn.length > 0) {
          findings.push({
            check: "Audience Expansion Enabled",
            severity: "critical",
            affectedEntities: expansionOn.map((c) => c.id),
            description: `${expansionOn.length} campaign(s) have audience expansion ON — LinkedIn will show ads to people outside your targeting criteria.`,
            recommendation:
              "Disable audience expansion immediately. It dilutes targeting precision and wastes budget on irrelevant impressions.",
          });
        }

        // ── Check 6: Audience network enabled ───────────────────
        const networkOn = campaigns.filter((c) => c.offsiteDeliveryEnabled);
        if (networkOn.length > 0) {
          findings.push({
            check: "Audience Network Enabled",
            severity: "critical",
            affectedEntities: networkOn.map((c) => c.id),
            description: `${networkOn.length} campaign(s) have LinkedIn Audience Network ON — ads are being served on third-party sites.`,
            recommendation:
              "Disable the Audience Network. Offsite placements have significantly lower engagement and conversion rates.",
          });
        }

        // ── Check 7: Stale campaigns ────────────────────────────
        const recentDays = 7;
        const recentStart = new Date();
        recentStart.setDate(recentStart.getDate() - recentDays);
        const staleCampaigns = campaigns.filter((c) => {
          const campaignData = analyticsElements.filter((a) =>
            String(a.pivotValue).includes(c.id)
          );
          const recentImpressions = campaignData
            .filter((a) => {
              const dateStr = a.dateRange as { start?: { year: number; month: number; day: number } } | undefined;
              if (!dateStr?.start) return false;
              const d = new Date(dateStr.start.year, dateStr.start.month - 1, dateStr.start.day);
              return d >= recentStart;
            })
            .reduce((sum, a) => sum + (Number(a.impressions) || 0), 0);
          return recentImpressions === 0;
        });
        if (staleCampaigns.length > 0) {
          findings.push({
            check: "Stale Campaigns",
            severity: "info",
            affectedEntities: staleCampaigns.map((c) => c.id),
            description: `${staleCampaigns.length} active campaign(s) have delivered zero impressions in the last ${recentDays} days.`,
            recommendation:
              "Pause or archive campaigns with no delivery. Check if targeting is too narrow, bids are too low, or budgets are exhausted.",
          });
        }

        // ── Check 8: Creative fatigue ───────────────────────────
        // Simplified: flag if overall CTR in the last week is < 70% of the first week
        if (weeklyCtrs.length >= 2) {
          const firstWeekCtr = weeklyCtrs[0];
          const lastWeekCtr = weeklyCtrs[weeklyCtrs.length - 1];
          if (firstWeekCtr > 0 && lastWeekCtr / firstWeekCtr < 0.7) {
            findings.push({
              check: "Creative Fatigue",
              severity: "warning",
              affectedEntities: ["account-wide"],
              description: `CTR has dropped ${((1 - lastWeekCtr / firstWeekCtr) * 100).toFixed(0)}% from the first week (${(firstWeekCtr * 100).toFixed(2)}% → ${(lastWeekCtr * 100).toFixed(2)}%).`,
              recommendation:
                "Rotate in fresh creatives. Audiences seeing the same ads repeatedly leads to banner blindness and declining engagement.",
            });
          }
        }

        // ── Check 9: Overlapping audiences ──────────────────────
        const targetingGroups = campaigns
          .filter((c) => c.targetingCriteria)
          .map((c) => ({
            id: c.id,
            targeting: JSON.stringify(c.targetingCriteria),
          }));
        const duplicates = findDuplicateTargeting(targetingGroups);
        if (duplicates.length > 0) {
          findings.push({
            check: "Overlapping Audiences",
            severity: "warning",
            affectedEntities: duplicates.flat(),
            description: `${duplicates.length} pair(s) of campaigns share identical targeting criteria, causing self-competition in the auction.`,
            recommendation:
              "Differentiate targeting between campaigns or consolidate them. Overlapping audiences drive up your own CPCs.",
          });
        }

        // ── Check 10: Missing UTM parameters ────────────────────
        // This would require pulling creatives — simplified check
        findings.push({
          check: "UTM Parameters",
          severity: "info",
          affectedEntities: ["requires-creative-scan"],
          description:
            "UTM parameter check requires scanning creative destination URLs. Use list_creatives to audit individually.",
          recommendation:
            "Ensure all destination URLs include utm_source=linkedin&utm_medium=paid_social&utm_campaign={campaign_name}.",
        });

        // ── Compile report ──────────────────────────────────────
        const report: AuditReport = {
          accountId: client.config.adAccountId,
          timestamp: new Date().toISOString(),
          totalFindings: findings.length,
          critical: findings.filter((f) => f.severity === "critical").length,
          warnings: findings.filter((f) => f.severity === "warning").length,
          info: findings.filter((f) => f.severity === "info").length,
          findings,
        };

        return report;
      },
    },

    check_single_audit: {
      description:
        "Run a single audit check by name. Useful for targeted investigation.",
      schema: z.object({
        check: z
          .enum([
            "declining_ctr",
            "budget_utilization",
            "conversion_tracking",
            "bid_strategy",
            "audience_expansion",
            "audience_network",
            "stale_campaigns",
            "creative_fatigue",
            "overlapping_audiences",
            "utm_parameters",
          ])
          .describe("Which audit check to run"),
      }),
      handler: async (args: { check: string }) => {
        // Delegate to the full audit but filter to the requested check
        return {
          message: `Run run_full_audit and filter for "${args.check}" — or use this as a targeted entry point.`,
          check: args.check,
        };
      },
    },
  };

src/tools/audit.ts:340-366 (helper)

Helper function computeWeeklyCtr that groups daily analytics data by week and computes CTR (clicks/impressions) for each week, used by the declining CTR and creative fatigue checks.

function computeWeeklyCtr(
  dailyData: Array<Record<string, unknown>>
): number[] {
  // Group by week and compute CTR per week
  const weeks: Map<number, { impressions: number; clicks: number }> = new Map();

  for (const row of dailyData) {
    const dateRange = row.dateRange as {
      start?: { year: number; month: number; day: number };
    } | undefined;
    if (!dateRange?.start) continue;

    const d = new Date(dateRange.start.year, dateRange.start.month - 1, dateRange.start.day);
    const weekNum = Math.floor(d.getTime() / (7 * 24 * 60 * 60 * 1000));

    const existing = weeks.get(weekNum) || { impressions: 0, clicks: 0 };
    existing.impressions += Number(row.impressions) || 0;
    existing.clicks += Number(row.clicks) || 0;
    weeks.set(weekNum, existing);
  }

  return [...weeks.entries()]
    .sort(([a], [b]) => a - b)
    .map(([, data]) =>
      data.impressions > 0 ? data.clicks / data.impressions : 0
    );
}

src/tools/audit.ts:368-373 (helper)

Helper function isConsistentlyDeclining that checks if an array of weekly CTR values is strictly decreasing, used by check 1 (Declining CTR).

function isConsistentlyDeclining(values: number[]): boolean {
  for (let i = 1; i < values.length; i++) {
    if (values[i] >= values[i - 1]) return false;
  }
  return true;
}

src/tools/audit.ts:375-387 (helper)

Helper function findDuplicateTargeting that compares JSON-stringified targeting criteria across campaigns to find overlapping audience targeting, used by check 9.

function findDuplicateTargeting(
  items: Array<{ id: string; targeting: string }>
): string[][] {
  const pairs: string[][] = [];
  for (let i = 0; i < items.length; i++) {
    for (let j = i + 1; j < items.length; j++) {
      if (items[i].targeting === items[j].targeting) {
        pairs.push([items[i].id, items[j].id]);
      }
    }
  }
  return pairs;
}

Tool Definition Quality

A3.5/5.0

Behavior2/5

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

No annotations are provided, so the description bears full responsibility for disclosing behavior. It states 'run' without clarifying if the tool modifies data, requires specific permissions, or produces a report. It lacks details on side effects, rate limits, or auth needs.

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?

A single sentence efficiently communicates the tool's purpose and the 10 checks it covers. No unnecessary words or redundancy.

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

Completeness3/5

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

Given no output schema and no annotations, the description adequately lists the audit areas but fails to describe the return format or how results are presented. This is a gap for an analysis tool.

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?

The schema already provides a description for the single parameter 'lookbackDays' (default 28). With 100% schema coverage, the description adds no additional semantic value beyond what the schema provides.

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

Purpose5/5

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

The description explicitly lists 10 specific audit checks (e.g., declining CTR, missing conversion tracking), clearly stating what the tool does. It distinguishes itself from the sibling 'check_single_audit' by implying a comprehensive, multi-check scope.

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?

While the description implies a comprehensive audit for campaign health, it provides no explicit guidance on when to use this tool versus the sibling 'check_single_audit' or any prerequisites. Usage context is implied but not clarified.

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

Install Server

Latest Blog Posts

Lightport: Open-Sourcing Glama's AI Gateway
By punkpeye on April 27, 2026.
open source
OpenAI
Tool Definition Quality Score (TDQS)
By punkpeye on April 3, 2026.
mcp
The Hackers Who Tracked My Sleep Cycle
By punkpeye on March 26, 2026.
security

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/charltonmediagroup/LICampaignsMCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server