| 1 | import type { CuratedDispatch, FeedItem, RepoStats } from "./types"; |
| 2 | |
| 3 | const FALLBACK_BASE = "https://api.deepseek.com"; |
| 4 | const FALLBACK_MODEL = "deepseek-flash"; |
| 5 | |
| 6 | interface ChatMessage { |
| 7 | role: "system" | "user" | "assistant"; |
| 8 | content: string; |
| 9 | } |
| 10 | |
| 11 | interface ChatResponse { |
| 12 | choices: { message: { content: string } }[]; |
| 13 | } |
| 14 | |
| 15 | export interface DeepSeekEnv { |
| 16 | baseUrl?: string; |
| 17 | model?: string; |
| 18 | } |
| 19 | |
| 20 | export async function chat( |
| 21 | messages: ChatMessage[], |
| 22 | apiKey: string, |
| 23 | jsonMode = false, |
| 24 | dsEnv?: DeepSeekEnv |
| 25 | ): Promise<string> { |
| 26 | const base = dsEnv?.baseUrl ?? process.env.DEEPSEEK_BASE_URL ?? FALLBACK_BASE; |
| 27 | const model = dsEnv?.model ?? process.env.DEEPSEEK_MODEL ?? FALLBACK_MODEL; |
| 28 | const res = await fetch(`${base}/v1/chat/completions`, { |
| 29 | method: "POST", |
| 30 | // Same bound as agentChat: a stalled provider must not hold the cron run. |
| 31 | signal: AbortSignal.timeout(180_000), |
| 32 | headers: { |
| 33 | "Content-Type": "application/json", |
| 34 | Authorization: `Bearer ${apiKey}`, |
| 35 | }, |
| 36 | body: JSON.stringify({ |
| 37 | model, |
| 38 | messages, |
| 39 | temperature: 0.4, |
| 40 | max_tokens: 4096, |
| 41 | reasoning_effort: "high", |
| 42 | ...(jsonMode ? { response_format: { type: "json_object" } } : {}), |
| 43 | }), |
| 44 | }); |
| 45 | if (!res.ok) { |
| 46 | const text = await res.text(); |
| 47 | throw new Error(`DeepSeek ${res.status}: ${text}`); |
| 48 | } |
| 49 | const data = (await res.json()) as ChatResponse; |
| 50 | return data.choices[0]?.message?.content ?? ""; |
| 51 | } |
| 52 | |
| 53 | const SYSTEM_PROMPT = `You are the editor of "今日要闻 / Today's Dispatch", a daily-ish digest for the Codewhale open source project. |
| 54 | |
| 55 | You receive: repo stats and a list of recently updated issues, PRs, and releases. |
| 56 | Output a single JSON object — no prose around it — matching this exact shape: |
| 57 | |
| 58 | { |
| 59 | "headline": "string — English editorial headline (max ~70 chars)", |
| 60 | "summary": "string — 2-3 English sentences, calm factual editorial voice", |
| 61 | "highlights": [ |
| 62 | { "title": "string", "href": "string", "tag": "shipped|merged|opened|discussion|release", "blurb": "one sentence, max ~120 chars" } |
| 63 | ], |
| 64 | "movers": [ |
| 65 | { "number": 123, "title": "string", "href": "string", "reason": "one short clause" } |
| 66 | ], |
| 67 | "headlineZh": "string — Chinese (zh-CN) editorial headline, rewritten natively, not translated", |
| 68 | "summaryZh": "string — 2-3 Chinese sentences, native zh-CN prose", |
| 69 | "highlightsZh": [ |
| 70 | { "title": "string — zh-CN native", "href": "string", "tag": "shipped|merged|opened|discussion|release", "blurb": "zh-CN one sentence" } |
| 71 | ], |
| 72 | "moversZh": [ |
| 73 | { "number": 123, "title": "string — zh-CN", "href": "string", "reason": "zh-CN clause" } |
| 74 | ] |
| 75 | } |
| 76 | |
| 77 | Rules: |
| 78 | - Pick 3-5 highlights and 3-5 movers from the actual provided items. Never invent. |
| 79 | - Prefer items with discussion, merged PRs, recent releases, or labelled "good first issue". |
| 80 | - Tone: like a small-paper editor — measured, specific, never breathless. |
| 81 | - Never use words like "exciting", "amazing", "powerful", "revolutionary" (or Chinese equivalents like 令人兴奋, 强大无比, 革命性). |
| 82 | - href must be the html_url provided. |
| 83 | - The zh-CN fields must be native Chinese prose — not a direct translation of the English fields. Write them as a Chinese-speaking maintainer would. |
| 84 | - zh-CN uses full-width punctuation in CJK sentences (。，、).`; |
| 85 | |
| 86 | export async function curate( |
| 87 | apiKey: string, |
| 88 | stats: RepoStats, |
| 89 | feed: FeedItem[], |
| 90 | dsEnv?: DeepSeekEnv |
| 91 | ): Promise<CuratedDispatch> { |
| 92 | const trimmedFeed = feed.slice(0, 25).map((f) => ({ |
| 93 | kind: f.kind, |
| 94 | number: f.number, |
| 95 | title: f.title, |
| 96 | state: f.state, |
| 97 | href: f.url, |
| 98 | author: f.author, |
| 99 | updated: f.updatedAt, |
| 100 | comments: f.comments, |
| 101 | labels: f.labels.map((l) => l.name), |
| 102 | })); |
| 103 | |
| 104 | const userPayload = { |
| 105 | repo: "codewhale-hq/CodeWhale", |
| 106 | stats: { |
| 107 | stars: stats.stars, |
| 108 | forks: stats.forks, |
| 109 | open_issues: stats.openIssues, |
| 110 | open_pulls: stats.openPulls, |
| 111 | latest_release: stats.latestRelease?.tag, |
| 112 | }, |
| 113 | recent: trimmedFeed, |
| 114 | }; |
| 115 | |
| 116 | const raw = await chat( |
| 117 | [ |
| 118 | { role: "system", content: SYSTEM_PROMPT }, |
| 119 | { role: "user", content: JSON.stringify(userPayload, null, 2) }, |
| 120 | ], |
| 121 | apiKey, |
| 122 | true, |
| 123 | dsEnv |
| 124 | ); |
| 125 | |
| 126 | const parsed = JSON.parse(raw) as Omit<CuratedDispatch, "generatedAt">; |
| 127 | return { ...sanitizeDispatch(parsed), generatedAt: new Date().toISOString() }; |
| 128 | } |
| 129 | |
| 130 | const SAFE_HREF_RE = /^https:\/\/(?:github\.com|api\.github\.com|codewhale\.net|crates\.io|www\.npmjs\.com|docs\.rs)\//; |
| 131 | const FALLBACK_HREF = "https://github.com/codewhale-hq/CodeWhale"; |
| 132 | |
| 133 | function safeHref(u: unknown): string { |
| 134 | return typeof u === "string" && SAFE_HREF_RE.test(u) ? u : FALLBACK_HREF; |
| 135 | } |
| 136 | |
| 137 | function sanitizeDispatch(d: Omit<CuratedDispatch, "generatedAt">): Omit<CuratedDispatch, "generatedAt"> { |
| 138 | return { |
| 139 | ...d, |
| 140 | highlights: (d.highlights ?? []).map((h) => ({ ...h, href: safeHref(h.href) })), |
| 141 | movers: (d.movers ?? []).map((m) => ({ ...m, href: safeHref(m.href) })), |
| 142 | highlightsZh: (d.highlightsZh ?? []).map((h) => ({ ...h, href: safeHref(h.href) })), |
| 143 | moversZh: (d.moversZh ?? []).map((m) => ({ ...m, href: safeHref(m.href) })), |
| 144 | }; |
| 145 | } |
| 146 |