@sellable/mcp 0.1.269 → 0.1.271

package/dist/tools/harvest-jobs.js

@@ -1,429 +0,0 @@
-import * as fs from "node:fs";
-import * as os from "node:os";
-import * as path from "node:path";
-import { getApi } from "../api.js";
-import { resolveWorkspaceRoot } from "../utils/workspace-root.js";
-const entryPath = process.argv[1] ? path.resolve(process.argv[1]) : process.cwd();
-const workspaceRoot = resolveWorkspaceRoot(path.dirname(entryPath));
-const HARVEST_JOB_SEARCH_TOKEN_REF_PREFIX = "mcp-harvest-job-search-token:";
-const MAX_HARVEST_JOB_SEARCH_TOKEN_REFS = 200;
-const tokenRefs = new Map();
-let tokenRefCounter = 0;
-export const harvestJobToolDefinitions = [
-    {
-        name: "search_harvest_jobs",
-        description: "Search current LinkedIn jobs through Sellable's Harvest-backed v3 API, write markdown/CSV review artifacts, and return a token reference for selected-company confirmation. Use when account-source intent is current LinkedIn job posts. Does not call Prospeo or confirm companies.",
-        inputSchema: {
-            type: "object",
-            properties: {
-                search: { type: "string" },
-                searches: { type: "array", items: { type: "string" } },
-                location: { type: "string" },
-                geoId: { type: "string" },
-                postedLimit: { type: "string", enum: ["24h", "week", "month"] },
-                sortBy: { type: "string", enum: ["relevance", "date"] },
-                workplaceType: {
-                    oneOf: [
-                        { type: "string" },
-                        { type: "array", items: { type: "string" } },
-                    ],
-                },
-                employmentType: {
-                    oneOf: [
-                        { type: "string" },
-                        { type: "array", items: { type: "string" } },
-                    ],
-                },
-                experienceLevel: {
-                    oneOf: [
-                        { type: "string" },
-                        { type: "array", items: { type: "string" } },
-                    ],
-                },
-                easyApply: { type: "boolean" },
-                under10Applicants: { type: "boolean" },
-                salary: { type: "string" },
-                pages: {
-                    type: "number",
-                    description: "Max pages per search term. Backend cap is 2.",
-                },
-                maxRows: {
-                    type: "number",
-                    description: "Max returned rows across searches. Backend cap is 100.",
-                },
-                artifactFormat: {
-                    type: "string",
-                    enum: ["markdown", "csv", "both"],
-                    description: "Review artifact format. Defaults to markdown.",
-                },
-                outputDir: {
-                    type: "string",
-                    description: "Optional safe output directory under the workspace, home, or temp root.",
-                },
-                fileBaseName: {
-                    type: "string",
-                    description: "Optional safe artifact filename base without extension.",
-                },
-            },
-            required: [],
-            additionalProperties: false,
-        },
-    },
-    {
-        name: "confirm_harvest_job_companies",
-        description: "Confirm selected Harvest LinkedIn job IDs into a backend-resolved Prospeo domainFilterId, then guide the follow-on search_prospeo people search. Requires searchToken or token reference from search_harvest_jobs. Does not accept raw domains.",
-        inputSchema: {
-            type: "object",
-            properties: {
-                searchToken: {
-                    type: "string",
-                    description: "Signed search token or mcp-harvest-job-search-token:* reference from search_harvest_jobs.",
-                },
-                selectedJobIds: {
-                    type: "array",
-                    items: { type: "string" },
-                    description: "Harvest job IDs selected from the search artifact.",
-                },
-                name: { type: "string" },
-                outputDir: {
-                    type: "string",
-                    description: "Optional safe output directory under the workspace, home, or temp root.",
-                },
-                fileBaseName: {
-                    type: "string",
-                    description: "Optional safe artifact filename base without extension.",
-                },
-            },
-            required: ["searchToken", "selectedJobIds"],
-            additionalProperties: false,
-        },
-    },
-];
-function removeUndefinedValues(input) {
-    return Object.fromEntries(Object.entries(input).filter(([, value]) => value !== undefined));
-}
-function storeTokenRef(token) {
-    if (typeof token !== "string" || token.length === 0)
-        return token;
-    if (token.startsWith(HARVEST_JOB_SEARCH_TOKEN_REF_PREFIX))
-        return token;
-    const ref = `${HARVEST_JOB_SEARCH_TOKEN_REF_PREFIX}${Date.now().toString(36)}-${(++tokenRefCounter).toString(36)}`;
-    tokenRefs.set(ref, token);
-    while (tokenRefs.size > MAX_HARVEST_JOB_SEARCH_TOKEN_REFS) {
-        const oldest = tokenRefs.keys().next().value;
-        if (!oldest)
-            break;
-        tokenRefs.delete(oldest);
-    }
-    return ref;
-}
-function resolveTokenRef(token) {
-    if (!token.startsWith(HARVEST_JOB_SEARCH_TOKEN_REF_PREFIX))
-        return token;
-    const resolved = tokenRefs.get(token);
-    if (!resolved) {
-        throw new Error("confirm_harvest_job_companies received a stale Harvest job search token reference. Re-run search_harvest_jobs.");
-    }
-    return resolved;
-}
-function validateSearchInput(input) {
-    if (!input?.search && (!input?.searches || input.searches.length === 0)) {
-        throw new Error("search_harvest_jobs requires search or searches.");
-    }
-}
-function validateConfirmInput(input) {
-    if (!input?.searchToken) {
-        throw new Error("confirm_harvest_job_companies requires searchToken from search_harvest_jobs.");
-    }
-    if (!input.selectedJobIds || input.selectedJobIds.length === 0) {
-        throw new Error("confirm_harvest_job_companies requires selectedJobIds from search_harvest_jobs results.");
-    }
-    if ("domain" in input ||
-        "domains" in input ||
-        "includeDomains" in input ||
-        "companyDomains" in input) {
-        throw new Error("confirm_harvest_job_companies does not accept raw domains. Use searchToken and selectedJobIds.");
-    }
-}
-function decodeMaybe(value) {
-    try {
-        return decodeURIComponent(value);
-    }
-    catch {
-        return value;
-    }
-}
-function isPathInside(candidate, root) {
-    const relative = path.relative(root, candidate);
-    return (relative === "" ||
-        (!relative.startsWith("..") && !path.isAbsolute(relative)));
-}
-function safeBaseName(value, fallback) {
-    const raw = (value ?? fallback).trim();
-    const decoded = decodeMaybe(raw);
-    const cleaned = decoded
-        .replace(/\.[A-Za-z0-9]+$/, "")
-        .replace(/[^A-Za-z0-9._-]+/g, "-")
-        .replace(/^-+|-+$/g, "")
-        .slice(0, 80);
-    if (!cleaned || cleaned.includes("..")) {
-        throw new Error("fileBaseName must be a safe filename.");
-    }
-    return cleaned;
-}
-function allowedRoots() {
-    return [workspaceRoot, os.homedir(), os.tmpdir()]
-        .map((root) => {
-        const resolved = path.resolve(root);
-        try {
-            return fs.realpathSync(resolved);
-        }
-        catch {
-            return resolved;
-        }
-    })
-        .filter(Boolean);
-}
-function resolveOutputDir(outputDir) {
-    const target = outputDir
-        ? path.isAbsolute(outputDir)
-            ? path.resolve(outputDir)
-            : path.resolve(workspaceRoot, outputDir)
-        : path.resolve(workspaceRoot, ".sellable", "artifacts", "harvest-jobs");
-    if (target.includes("\0")) {
-        throw new Error("outputDir contains invalid characters.");
-    }
-    fs.mkdirSync(target, { recursive: true });
-    const real = fs.realpathSync(target);
-    if (!allowedRoots().some((root) => isPathInside(real, root))) {
-        throw new Error("outputDir must be inside the workspace, home directory, or temp directory.");
-    }
-    return real;
-}
-function collisionSafePath(outputDir, fileBaseName, ext) {
-    const extension = ext.startsWith(".") ? ext : `.${ext}`;
-    if (![".md", ".csv"].includes(extension)) {
-        throw new Error(`Unsupported artifact extension: ${extension}`);
-    }
-    let candidate = path.join(outputDir, `${fileBaseName}${extension}`);
-    let counter = 1;
-    while (fs.existsSync(candidate)) {
-        candidate = path.join(outputDir, `${fileBaseName}-${counter}${extension}`);
-        counter += 1;
-    }
-    const resolved = path.resolve(candidate);
-    if (!isPathInside(resolved, outputDir)) {
-        throw new Error("Artifact path escapes outputDir.");
-    }
-    if (fs.existsSync(resolved) && fs.lstatSync(resolved).isSymbolicLink()) {
-        throw new Error("Refusing to write through symlink artifact path.");
-    }
-    return resolved;
-}
-function csvEscape(value) {
-    const text = value === null || value === undefined ? "" : String(value);
-    return /[",\n\r]/.test(text) ? `"${text.replace(/"/g, '""')}"` : text;
-}
-function formatValue(value) {
-    if (value === null || value === undefined)
-        return "";
-    if (typeof value === "object")
-        return JSON.stringify(value);
-    return String(value);
-}
-function renderSearchMarkdown(response) {
-    const rows = Array.isArray(response.rows) ? response.rows : [];
-    const lines = [
-        "# Harvest LinkedIn Job Search",
-        "",
-        `Row count: ${response.rowCount ?? rows.length}`,
-        `Cost summary: ${JSON.stringify(response.costSummary ?? {})}`,
-        `Search summary: ${JSON.stringify(response.summary ?? {})}`,
-        "",
-        "Domains are resolved only during confirmation. Select stable Harvest job IDs, not LinkedIn company URLs.",
-        "",
-        "| Job ID | Company | Title | Posted | Location | Easy Apply | Job URL | Company LinkedIn | Request ID |",
-        "| --- | --- | --- | --- | --- | --- | --- | --- | --- |",
-        ...rows.map((row) => [
-            row.jobId,
-            row.companyName,
-            row.title,
-            row.postedDate,
-            row.locationText,
-            row.easyApply,
-            row.jobUrl,
-            row.companyLinkedinUrl,
-            row.harvestRequestId,
-        ]
-            .map((value) => String(value ?? "").replace(/\|/g, "\\|"))
-            .join(" | ")
-            .replace(/^/, "| ")
-            .replace(/$/, " |")),
-        "",
-        "Next step: confirm_harvest_job_companies({ searchToken, selectedJobIds: [...] })",
-    ];
-    return `${lines.join("\n")}\n`;
-}
-function renderSearchCsv(response) {
-    const rows = Array.isArray(response.rows) ? response.rows : [];
-    const columns = [
-        "rowId",
-        "jobId",
-        "companyName",
-        "title",
-        "jobUrl",
-        "postedDate",
-        "locationText",
-        "easyApply",
-        "companyLinkedinUrl",
-        "companyUniversalName",
-        "harvestRequestId",
-        "search",
-        "page",
-    ];
-    return [
-        columns.join(","),
-        ...rows.map((row) => columns.map((column) => csvEscape(row[column])).join(",")),
-    ].join("\n");
-}
-function renderConfirmMarkdown(response) {
-    const unresolved = Array.isArray(response.unresolvedJobs)
-        ? response.unresolvedJobs
-        : [];
-    const skipped = Array.isArray(response.skippedJobs) ? response.skippedJobs : [];
-    const lines = [
-        "# Harvest LinkedIn Job Confirmation",
-        "",
-        `Domain filter ID: ${response.domainFilterId ?? ""}`,
-        `Resolved: ${response.resolvedCount ?? 0}`,
-        `Unresolved: ${response.unresolvedCount ?? 0}`,
-        `Duplicates: ${response.duplicateCount ?? 0}`,
-        `Skipped: ${response.skippedCount ?? 0}`,
-        `Cost summary: ${JSON.stringify(response.costSummary ?? {})}`,
-        "",
-        "## Resolved Domains",
-        "",
-        ...(response.includeDomains ?? []).map((domain) => `- ${domain}`),
-        "",
-        "## Unresolved Jobs",
-        "",
-        ...unresolved.map((row) => `- ${formatValue(row.jobId)}: ${formatValue(row.reason)}`),
-        "",
-        "## Skipped Jobs",
-        "",
-        ...skipped.map((row) => `- ${formatValue(row.jobId)}: ${formatValue(row.reason)}`),
-        "",
-        "Next step: search_prospeo with the returned domainFilterId.",
-    ];
-    return `${lines.join("\n")}\n`;
-}
-function writeSearchArtifacts(response, input) {
-    const format = input.artifactFormat ?? "markdown";
-    const outputDir = resolveOutputDir(input.outputDir);
-    const baseName = safeBaseName(input.fileBaseName, `harvest-jobs-${new Date().toISOString().replace(/[:.]/g, "-")}`);
-    const artifacts = {};
-    if (format === "markdown" || format === "both") {
-        const markdownPath = collisionSafePath(outputDir, baseName, ".md");
-        fs.writeFileSync(markdownPath, renderSearchMarkdown(response), "utf8");
-        artifacts.markdown = markdownPath;
-    }
-    if (format === "csv" || format === "both") {
-        const csvPath = collisionSafePath(outputDir, baseName, ".csv");
-        fs.writeFileSync(csvPath, renderSearchCsv(response), "utf8");
-        artifacts.csv = csvPath;
-    }
-    return artifacts;
-}
-function writeConfirmArtifact(response, input) {
-    const outputDir = resolveOutputDir(input.outputDir);
-    const baseName = safeBaseName(input.fileBaseName, `harvest-jobs-confirm-${new Date().toISOString().replace(/[:.]/g, "-")}`);
-    const artifactPath = collisionSafePath(outputDir, baseName, ".md");
-    fs.writeFileSync(artifactPath, renderConfirmMarkdown(response), "utf8");
-    return artifactPath;
-}
-function compactSearchResponse(response, artifacts) {
-    const rows = Array.isArray(response.rows) ? response.rows : [];
-    const token = storeTokenRef(response.searchToken);
-    return removeUndefinedValues({
-        success: response.success ?? true,
-        rowCount: typeof response.rowCount === "number" ? response.rowCount : rows.length,
-        costSummary: response.costSummary ?? {},
-        artifacts,
-        sampleJobs: rows.slice(0, 10).map((row) => removeUndefinedValues({
-            rowId: row.rowId,
-            jobId: row.jobId,
-            company: row.companyName,
-            title: row.title,
-            jobUrl: row.jobUrl,
-            postedDate: row.postedDate,
-            location: row.locationText,
-            easyApply: row.easyApply,
-            companyLinkedinUrl: row.companyLinkedinUrl,
-            companyUniversalName: row.companyUniversalName,
-        })),
-        searchToken: token,
-        nextStep: "Review the artifact, then call confirm_harvest_job_companies with searchToken and selectedJobIds. Harvest jobs are account-source evidence; Prospeo is the follow-on people search.",
-    });
-}
-function compactConfirmResponse(response, artifactPath) {
-    return removeUndefinedValues({
-        success: response.success ?? true,
-        domainFilterId: response.domainFilterId,
-        includeDomains: response.includeDomains,
-        resolvedCount: response.resolvedCount,
-        unresolvedCount: response.unresolvedCount,
-        duplicateCount: response.duplicateCount,
-        skippedCount: response.skippedCount,
-        costSummary: response.costSummary,
-        unresolvedJobs: response.unresolvedJobs ?? [],
-        skippedJobs: response.skippedJobs ?? [],
-        carryoverSummary: response.carryoverSummary,
-        confirmArtifactPath: artifactPath,
-        suggestedNextCall: {
-            tool: "search_prospeo",
-            arguments: response.nextSearchProspeoCall ?? {
-                domainFilterId: response.domainFilterId,
-                filters: { max_person_per_company: 1 },
-            },
-        },
-    });
-}
-export async function searchHarvestJobs(input) {
-    validateSearchInput(input);
-    const api = getApi();
-    const response = await api.post("/api/v3/harvest/linkedin-jobs/search", removeUndefinedValues({
-        search: input.search,
-        searches: input.searches,
-        location: input.location,
-        geoId: input.geoId,
-        postedLimit: input.postedLimit,
-        sortBy: input.sortBy,
-        workplaceType: input.workplaceType,
-        employmentType: input.employmentType,
-        experienceLevel: input.experienceLevel,
-        easyApply: input.easyApply,
-        under10Applicants: input.under10Applicants,
-        salary: input.salary,
-        pages: input.pages,
-        maxRows: input.maxRows,
-    }));
-    if (!response || typeof response !== "object" || !response.searchToken) {
-        throw new Error("search_harvest_jobs API response did not include a searchToken.");
-    }
-    const artifacts = writeSearchArtifacts(response, input);
-    return compactSearchResponse(response, artifacts);
-}
-export async function confirmHarvestJobCompanies(input) {
-    validateConfirmInput(input);
-    const api = getApi();
-    const response = await api.post("/api/v3/harvest/linkedin-jobs/confirm-domain-filter", removeUndefinedValues({
-        searchToken: resolveTokenRef(input.searchToken),
-        selectedJobIds: input.selectedJobIds,
-        name: input.name,
-    }));
-    if (!response || typeof response !== "object" || !response.domainFilterId) {
-        throw new Error("confirm_harvest_job_companies API response did not include a domainFilterId.");
-    }
-    const artifactPath = writeConfirmArtifact(response, input);
-    return compactConfirmResponse(response, artifactPath);
-}