Redbook — Xiaohongshu CLI

Use the redbook CLI to search notes, read content, analyze creators, automate engagement, and research topics on Xiaohongshu (小红书/RED).

OpenClaw users: Install via clawhub install redbook or npm install -g @lucasygu/redbook.

Usage

CODEBLOCK0

Quick Reference

Intent	Command
Search notes	INLINECODE3
Read a note

redbook read <url> --json | | Get comments | redbook comments <url> --json --all | | Creator profile | redbook user <userId> --json | | Creator's posts | redbook user-posts <userId> --json | | Browse feed | redbook feed --json | | Search hashtags | redbook topics "keyword" --json | | Analyze viral note | redbook analyze-viral <url> --json | | Extract content template | redbook viral-template <url1> <url2> --json | | Post a comment | redbook comment <url> --content "text" | | Reply to comment | redbook reply <url> --comment-id <id> --content "text" | | Batch reply (preview) | redbook batch-reply <url> --strategy questions --dry-run | | Like a note | redbook like <url> | | Unlike a note | redbook like <url> --undo | | List favorites | redbook favorites --json or redbook favorites <userId> --json | | Collect a note | redbook collect <url> | | Remove from collection | redbook uncollect <url> | | List followers | redbook followers <userId> --json | | List following | redbook following <userId> --json | | Delete own note | redbook delete <url> | | Check note health | redbook health --json or redbook health --all --json | | List user boards | redbook boards or redbook boards <userId> --json | | List album notes | redbook board <board-url> or redbook board <boardId> --json | | Render markdown to cards | redbook render content.md --style xiaohongshu | | Publish image note | redbook post --title "..." --body "..." --images img.jpg | | Check connection | redbook whoami |

Always add --json when parsing output programmatically. Without it, output is human-formatted text.

---

XHS Platform Signals

XHS is not Twitter or Instagram. These platform-specific engagement ratios reveal content type and audience behavior.

Collect/Like Ratio (collected_count / liked_count)

XHS's "collect" (收藏) is a save-for-later mechanic — users build personal reference libraries. This ratio is the strongest signal of content utility.

Ratio	Classification	Meaning
>40%	工具型 (Reference)	Tutorial, checklist, template — users bookmark for reuse
20–40%

认知型 (Insight) | Thought-provoking but not saved for later |
| <20% | 娱乐型 (Entertainment) | Consumed and forgotten — engagement is passive |

Comment/Like Ratio (comment_count / liked_count)

Measures how much a note triggers conversation.

Ratio	Classification	Meaning
>15%	讨论型 (Discussion)	Debate, sharing experiences, asking questions
5–15%

正常互动 (Normal) | Typical engagement pattern |
| <5% | 围观型 (Passive) | Users like but don't engage further |

Share/Like Ratio (share_count / liked_count)

Measures social currency — whether users share to signal identity or help others.

Ratio	Meaning
>10%	社交货币 — people share to signal taste, identity, or help friends
<10%

Content consumed individually, not forwarded |

Search Sort Semantics

Sort	What It Reveals
INLINECODE37	Proven ceiling — the best a keyword can do
INLINECODE38

Content velocity — how much is being posted now | | --sort general | Algorithm-weighted blend (default) |

Content Form Dynamics

Form	Tendency
图文 (image-text, type: "normal")	Higher collect rate — users save reference content
视频 (video, type: "video")

Higher like rate — easier to consume passively |

---

Analysis Modules

Each module is a composable building block. Combine them for different analysis depths.

Module A: Keyword Engagement Matrix

Answers: Which keywords have the highest engagement ceiling? Which are saturated vs. underserved?

Commands:
CODEBLOCK1

Fields to extract from each result's items[]:

• - items[].note_card.interact_info.liked_count — likes (may use Chinese numbers: "1.5万" = 15,000)
• INLINECODE44 — collects
• INLINECODE45 — comments
• INLINECODE46 — author

How to interpret:

• - Top1 ceiling = items[0] likes — the best-performing note for this keyword. This is the proven demand signal.
• Top10 average = mean likes across items[0..9] — how well an average top note does.
• A high Top1 but low Top10 avg means one outlier dominates; hard to compete.
• A high Top10 avg means consistent demand; easier to break in.

Output: Keyword × engagement table ranked by Top1 ceiling.

Keyword	Top1 Likes	Top10 Avg	Top1 Collects	Collect/Like
keyword1	12,000	3,200	5,400	45%
keyword2

8,500 | 4,100 | 1,200 | 14% |

---

Module B: Cross-Topic Heatmap

Answers: Which topic × scene intersections have demand? Where are the content gaps?

Commands:
CODEBLOCK2

Fields to extract: Same as Module A — Top1 liked_count for each combination.

How to interpret:

• - High Top1 = proven demand for this intersection
• Zero or very low results = content gap (opportunity or no demand — check if the combination makes sense)
• Compare across scenes to find which angles resonate most with the base topic

Output: Base × Scene heatmap.

CODEBLOCK3

---

Module C: Engagement Signal Analysis

Answers: What type of content is each keyword? Reference, insight, or entertainment?

Commands: Use search results from Module A, or for a single note:
CODEBLOCK4

Fields to extract:

• - From search results: compute ratios from interact_info fields
• From analyze-viral: use pre-computed engagement.collectToLikeRatio, engagement.commentToLikeRatio, INLINECODE54

How to interpret: Apply the ratio benchmarks from XHS Platform Signals above.

Output: Per-keyword or per-note classification.

Keyword	Collect/Like	Comment/Like	Type
keyword1	45%	8%	工具型 + 正常互动
keyword2

12% | 22% | 娱乐型 + 讨论型 |

---

Module D: Creator Discovery & Profiling

Answers: Who are the key creators in this niche? What are their strategies?

Commands:
CODEBLOCK5

Fields to extract:

• - From user: interactions[] where type === "fans" → follower count
• From user-posts: notes[].interact_info.liked_count for all posts → compute avg, median, max
• From user-posts: notes[].display_title → content patterns, posting frequency

How to interpret:

• - Avg vs. Median likes: Large gap means viral outliers inflate the average. Median is the "true" baseline.
• Max / Median ratio: >5× means they've had breakout hits. Study those notes specifically.
• Post frequency: Count notes to estimate posting cadence. Prolific creators (>3/week) vs. quality-focused (<1/week).

Output: Creator comparison table.

Creator	Followers	Avg Likes	Median	Max	Posts	Style
@creator1	12万	3,200	1,800	45,000	89	Tutorial
@creator2

5.4万 | 8,100 | 6,500 | 22,000 | 34 | Story |

---

Module E: Content Form Breakdown

Answers: Do image-text or video notes perform better for this topic?

Commands:
CODEBLOCK6

Fields to extract:

• - Compare Top1 and Top10 avg liked_count and collected_count between the two result sets
• Note the type field: "normal" = image-text, "video" = video

Output: Form × engagement table.

Form	Top1 Likes	Top10 Avg	Collect/Like
图文	8,000	2,400	42%
视频

15,000 | 5,100 | 18% |

---

Module F: Opportunity Scoring

Answers: Which keywords should I target? Where is the best effort-to-reward ratio?

Input: Keyword matrix from Module A.

Scoring logic:

• - Demand = Top1 likes ceiling (proven audience size)
• Competition = density of high-engagement results (how many notes in Top10 have >1K likes)
• Score = Demand × (1 / Competition density)

Tier thresholds (based on Top1 likes):

Tier	Top1 Likes	Meaning
S	>100,000 (10万+)	Massive demand — hard to compete but huge upside
A

20,000–100,000 | Strong demand — competitive but winnable |
| B | 5,000–20,000 | Moderate demand — good for growing accounts |
| C | <5,000 | Niche — low competition, low ceiling |

Output: Tiered keyword list.

Tier	Keyword	Top1	Competition	Opportunity
A	keyword1	45K	Medium (6/10 >1K)	High
B

keyword3 | 12K | Low (2/10 >1K) | Very High |
| S | keyword2 | 120K | High (10/10 >1K) | Medium |

---

Module G: Audience Inference

Answers: Who is the audience for this niche? What do they want?

Input: Engagement ratios from Module C + comment themes from analyze-viral + content patterns.

Fields to extract from analyze-viral JSON:

• - comments.themes[] — recurring phrases and keywords from comment section
• INLINECODE70 — % of comments that are questions (learning intent)
• INLINECODE71 — save behavior signals intent
• INLINECODE72 — what title patterns attract this audience

Inference rules:

• - High collect rate + high question rate → learning-oriented audience (students, professionals)
• High comment rate + emotional themes → community-oriented audience (sharing experiences)
• High share rate → aspiration-oriented audience (lifestyle, identity signaling)
• Comment language patterns → age/education signals (formal = older, slang = younger)

Output: Audience persona summary — demographics, intent, content preferences.

---

Module H: Content Brainstorm

Answers: What specific content should I create, backed by data?

Input: Opportunity scores (Module F) + audience persona (Module G) + heatmap gaps (Module B).

For each content idea, specify:

• - Target keyword — from opportunity scoring
• Hook angle — based on hookPatterns that work for this niche
• Content type — 工具型/认知型/娱乐型 based on what the audience wants
• Form — 图文 or 视频 based on Module E
• Engagement target — realistic based on Top10 avg for this keyword
• Competitive reference — specific note URL that proves this angle works

Output: Ranked content ideas with data backing.

#	Keyword	Hook Angle	Type	Target Likes	Reference
1	keyword3	"N个方法..." (List)	工具型 图文	5K+	[top note URL]
2

keyword1 | "为什么..." (Question) | 认知型 视频 | 10K+ | [top note URL] |

---

Module I: Comment Operations

Answers: Which comments deserve a reply? What is the comment quality distribution?

Commands:
CODEBLOCK7

Fields to extract from --dry-run JSON:

• - candidates[].commentId — target comments
• INLINECODE76 — boolean, detected question
• INLINECODE77 — engagement signal
• INLINECODE78 — whether already answered
• INLINECODE79 — how many comments were filtered out
• INLINECODE80 — total fetched

Strategies:

• - questions — replies to comments ending with ？ or ? (learning-oriented audience)
• INLINECODE84 — replies to highest-liked comments (maximum visibility)
• INLINECODE85 — replies to comments with no existing sub-replies (fill gaps)

How to interpret:

• - High question rate (>15%) = audience is learning-oriented → reply to build authority
• High top-engaged comments (>100 likes) = reply to visible comments for maximum reach
• Many unanswered comments = engagement gap, opportunity to increase reply rate

Safety: Hard cap 30 replies per batch, minimum 3-minute delay with ±30% jitter (default 5 min), --dry-run by default (no template = preview only), immediate stop on captcha. See Rate Limits & Safety for details.

Output: Reply plan table with candidate comments, strategy match reason, and status.

---

Module J: Viral Replication

Answers: What structural template can I extract from successful notes to guide new content creation?

Commands:
CODEBLOCK8

Fields to extract from viral-template JSON:

• - dominantHookPatterns[] — hook types appearing in majority of notes
• INLINECODE89 — specific title formula
• INLINECODE90 — target title length
• INLINECODE91 — target word count [min, max]
• INLINECODE92 — target paragraph count
• INLINECODE93 — reference/insight/entertainment
• INLINECODE94 — what the audience talks about

How to interpret:

• - Consistent hook patterns across notes = proven formula for this niche
• Narrow body length range = audience has clear content length preference
• High collect/like in profile = audience saves content → create reference material
• Common comment themes = topics to address in new content

Composition with other modules:

• - Uses Module A results to identify top URLs for template extraction
• Feeds into Module H (Content Brainstorm) as structural constraint
• Uses Module C classification to validate engagement profile

Output: Content template spec — the structural skeleton for content creation. An LLM (via the composed workflow) uses this template to generate actual title, body, hashtags, and cover image prompt.

---

Module K: Engagement Automation

Answers: How should I manage ongoing engagement with my audience?

This module is a workflow that composes Modules I and J with human oversight.

Workflow:

1. 1. Monitor — redbook comments "<myNoteUrl>" --all --json to fetch recent comments
2. Filter — redbook batch-reply --strategy questions --dry-run to identify reply candidates
3. Review — Human reviews dry-run output (or LLM reviews with persona guidelines)
4. Execute — INLINECODE97
5. Report — Summary of replies sent, errors encountered, rate limit status

Safety rules:

• - Always --dry-run first, human approval before execution
• Maximum 30 replies per session (hard cap)
• Minimum 3-minute delay between replies, default 5 minutes, with ±30% random jitter
• Never reply to the same comment twice (check hasSubReplies)
• Stop immediately on captcha — do not retry
• See Rate Limits & Safety for XHS risk control thresholds

Anti-spam guidelines:

• - Vary reply templates across batches
• Limit to 1-2 batch runs per note per day
• Prioritize quality (targeted strategy) over quantity
• Uniform timing patterns trigger bot detection — jitter is applied automatically

---

Module L: Card Rendering

Answers: How do I turn markdown content into Xiaohongshu-ready image cards?

Commands:
CODEBLOCK9

Input: Markdown file with YAML frontmatter:
CODEBLOCK10

Output: cover.png + card_1.png, card_2.png, ... in the same directory.

Card specs:

• - Size: 1080×1440 (3:4 ratio, standard XHS image)
• DPR: 2 (retina quality, actual output 2160×2880)
• Styles: purple, xiaohongshu, mint, sunset, ocean, elegant, dark

Pagination modes:

• - auto (default) — smart split on heading/paragraph boundaries using character-count heuristic
• INLINECODE104 — manual split on --- in markdown

How to interpret:

• - Uses the user's existing Chrome for rendering (via puppeteer-core) — no browser download needed
• Purely offline — no XHS API or cookies required
• Output images are ready for INLINECODE107

Dependencies: Requires puppeteer-core and marked (optional, install with npm install -g puppeteer-core marked).

Composition with other modules:

• - Pairs with Module H (Content Brainstorm) — generate content ideas, write markdown, render to cards
• Pairs with Module J (Viral Replication) — extract template, write content matching the template, render
• Output feeds into redbook post --images for publishing

---

Module M: Note Health Check (限流检测)

Answers: Are any of my notes being secretly rate-limited by XHS?

XHS assigns a hidden level field to each note in the creator backend API. This level controls recommendation distribution but is never shown in the UI. Your note may look "normal" while secretly receiving zero recommendations.

Commands:
CODEBLOCK11

Level definitions:

Level	Status	Meaning
4	🟢 Normal	Full recommendation distribution
2-3

🟡 Baseline | Basically normal, minor constraints |
| 1 | ⚪ New | Under review (new post) |
| -1 | 🔴 Soft limit | Mild throttling, decreased recommendations |
| -5 to -101 | 🔴 Moderate | Moderate throttling, minimal promotion |
| -102 | ⛔ Severe | Irreversible — must delete and repost |

Additional checks:

• - Sensitive word detection — flags titles containing automation/AI keywords (自动化, AI生成, 批量, etc.)
• Tag count warning — flags notes with >5 hashtags (risk factor)

How to interpret:

• - Level -1 or below = your note is being throttled. Consider editing or deleting + reposting.
• Level -102 = irreversible. Delete the note and create fresh content.
• Sensitive word hits = risky title keywords that may trigger throttling. Rephrase.
• Excessive tags = potential spam signal. Use 3-5 targeted tags.

Output: Terminal dashboard with color-coded distribution summary, limited notes list, and risk factor warnings.

Discovery credit: @xxx111god — xhs-note-health-checker

---

Composed Workflows

Combine modules for different analysis depths.

Quick Topic Scan (~5 min)

Modules: A → C → F

Search 3–5 keywords, classify engagement type, rank opportunities. Good for quickly validating whether a niche is worth deeper research.

Content Planning

Modules: A → B → E → F → H

Build keyword matrix, map topic × scene intersections, check content form performance, score opportunities, brainstorm specific content ideas.

Creator Competitive Analysis

Modules: A → D

Find who dominates a niche and study their content strategy, posting frequency, and engagement patterns.

Full Niche Analysis

Modules: A → B → C → D → E → F → G → H

The comprehensive playbook — keyword landscape, cross-topic heatmap, engagement signals, creator profiles, content form analysis, opportunity scoring, audience personas, and data-backed content ideas.

Single Note Deep-Dive

Command: INLINECODE113

No module composition needed — analyze-viral returns hook analysis, engagement ratios, comment themes, author baseline comparison, and a 0-100 viral score in one call.

Viral Pattern Research → Content Template

CODEBLOCK12

INLINECODE115 automates what previously required manual synthesis across analyze-viral results. It outputs a ContentTemplate JSON that captures dominant hooks, body structure ranges, engagement profile, and audience signals.

Reply Management

Modules: I

Single-module workflow for managing comment engagement on your notes. Use batch-reply --dry-run to audit, then execute with a template.

Content Replication

Modules: A → J → H → L

Keyword research → viral template extraction → data-backed content brainstorm → render to image cards. The template provides structural constraints that guide Module H's content ideas. Module L renders the final markdown to XHS-ready PNGs.

Content Creation End-to-End

Modules: A → J → H → L → INLINECODE119

The full pipeline: research keywords → extract viral template → brainstorm content → write markdown → render to styled image cards → publish via INLINECODE120

Account Health Monitoring

Modules: M

Run redbook health --all periodically to catch throttled notes early. If level drops below 1, investigate the note's content for policy violations. Combine with Module I to check if throttled notes still have unanswered comments worth replying to.

Full Operations

Modules: A → C → I → J → K → M

Comprehensive automation playbook — keyword analysis, engagement classification, comment operations, viral replication templates, and engagement automation workflow.

---

Rate Limits & Safety

XHS enforces aggressive anti-spam (风控) that detects automated behavior through device fingerprinting, activity ratio monitoring, and timing pattern analysis. The CLI applies safe defaults based on platform research.

Safe Thresholds

Action	Safe Interval	CLI Default	Hard Cap
Post a note	3-4 hours (2-3 notes/day max)	N/A (manual)	—
Comment

≥3 minutes | N/A (manual) | — | | Reply | ≥3 minutes | N/A (manual) | — | | Batch reply delay | ≥3 minutes | 5 min ±30% jitter | — | | Batch reply count | — | 10 | 30 |

Anti-Detection Measures

• - Timing jitter: ±30% random variation on all batch delays. Uniform intervals are a bot signature.
• Hard caps: Maximum 30 replies per batch (down from 50). No override.
• Rate limit warnings: post, comment, and reply commands display safe interval reminders after each action.
• Captcha circuit breaker: Batch operations stop immediately on captcha (NeedVerify).

What Triggers Risk Control

• - Uniform timing — replying at exact 3-second intervals flags bot detection
• High frequency — >50 interactions/minute across any action type
• Activity ratio anomaly — more comments than post views signals inauthentic behavior
• Device fingerprint mismatch — XHS fingerprints 21 hardware parameters

Best Practices for Agents

1. 1. Always --dry-run first, review candidates, then execute
2. Use the default 5-minute delay — do not override --delay below 180000 (3 min)
3. Limit batch runs to 1-2 per note per day
4. Vary reply templates between batches
5. Space post commands 3-4 hours apart (2-3 notes/day maximum)

---

API vs Browser Limitations

The following operations work reliably via API:

• - Reading: search, notes, comments, user profiles, feed, favorites
• Writing: top-level comments, comment replies, collect/uncollect notes
• Analysis: viral scoring, template extraction, batch reply planning

The following operations are unreliable via API (frequently trigger captcha):

• - Publishing notes (use --private for higher success rate)
• Bulk operations at very high frequency

The following operations require browser automation (not supported by this CLI):

• - Captcha solving, real-time notifications
• Like/follow (heavy anti-automation enforcement)
• DM/private messaging
• Cover image generation (use external tools like Gemini/DALL-E)

---

Command Details

redbook search <keyword>

Search for notes by keyword. Returns note titles, URLs, likes, author info.

CODEBLOCK13

Options:

• - --sort <type>: general (default), popular, INLINECODE133
• INLINECODE134: all (default), video, INLINECODE137
• INLINECODE138: Page number (default: 1)

redbook read <url>

Read a note's full content — title, body text, images, likes, comments count.

CODEBLOCK14

Accepts full URLs or short note IDs. Falls back to HTML scraping if API returns captcha.

redbook comments <url>

Get comments on a note. Use --all to fetch all pages.

CODEBLOCK15

redbook user <userId>

Get a creator's profile — nickname, bio, follower count, note count, likes received.

CODEBLOCK16

The userId is the hex string from the creator's profile URL.

redbook user-posts <userId>

List all notes posted by a creator. Returns titles, URLs, likes, timestamps.

CODEBLOCK17

redbook feed

Browse the recommendation feed.

CODEBLOCK18

redbook topics <keyword>

Search for topic hashtags. Useful for finding trending topics to attach to posts.

CODEBLOCK19

redbook favorites [userId]

List a user's collected (bookmarked) notes. Defaults to the current logged-in user when no userId is provided.

CODEBLOCK20

Options:

• - --all: Fetch all pages of favorites (default: first page only)

Note: Other users' favorites are only visible if they haven't set their collection to private.

redbook collect <url>

Collect (bookmark) a note to your favorites.

CODEBLOCK21

redbook uncollect <url>

Remove a note from your collection.

CODEBLOCK22

redbook analyze-viral <url>

Analyze why a viral note works. Returns a deterministic viral score (0–100).

CODEBLOCK23

Options:

• - --comment-pages <n>: Comment pages to fetch (default: 3, max: 10)

JSON output structure:
Returns { note, score, hook, content, visual, engagement, comments, relative, fetchedAt }.

• - score.overall (0–100) — composite of hook (20) + engagement (20) + relative (20) + content (20) + comments (20)
• INLINECODE154 — detected title patterns (Identity Hook, Emotion Word, Number Hook, Question, etc.)
• INLINECODE155 — likes, comments, collects, shares + ratios (collectToLikeRatio, commentToLikeRatio, shareToLikeRatio)
• INLINECODE156 — this note's likes / author's median likes
• INLINECODE157 — true if viralMultiplier > 3
• INLINECODE158 — top recurring keyword phrases from comments

redbook viral-template <url> [url2] [url3]

Extract a reusable content template from 1-3 viral notes. Analyzes each note (same pipeline as analyze-viral) and synthesizes common structural patterns.

CODEBLOCK24

Options:

• - --comment-pages <n>: Comment pages to fetch per note (default: 3, max: 10)

JSON output structure:
Returns { dominantHookPatterns, titleStructure, bodyStructure, engagementProfile, audienceSignals, sourceNotes, generatedAt }.

• - dominantHookPatterns[] — hook types appearing in majority of input notes
• INLINECODE164 — average title length across notes
• INLINECODE165 — [min, max] body length
• INLINECODE166 — "reference" / "insight" / "entertainment"
• INLINECODE167 — merged comment themes across notes

redbook comment <url>

Post a top-level comment on a note.

CODEBLOCK25

Options:

• - --content <text> (required): Comment text

redbook reply <url>

Reply to a specific comment on a note.

CODEBLOCK26

Options:

• - --comment-id <id> (required): Comment ID to reply to (from comments --json output)
• INLINECODE173 (required): Reply text

redbook batch-reply <url>

Reply to multiple comments using a filtering strategy. Always preview with --dry-run first.

CODEBLOCK27

Options:

• - --strategy <name>: questions (default), top-engaged, INLINECODE179
• INLINECODE180: Reply template with {author}, {content} placeholders
• INLINECODE183: Max replies (default: 10, hard cap: 30)
• INLINECODE184: Delay between replies in ms (default: 300000 / 5 min, min: 180000 / 3 min). ±30% random jitter applied automatically.
• INLINECODE185: Preview candidates without posting (default when no template)

Safety: Stops immediately on captcha. No template = dry-run only. Delays include random jitter to avoid uniform timing patterns that trigger XHS bot detection.

redbook render <file>

Render a markdown file with YAML frontmatter into styled PNG image cards. Uses the user's existing Chrome installation — no browser download needed.

CODEBLOCK28

Options:

• - --style <name>: purple, xiaohongshu (default), mint, sunset, ocean, elegant, INLINECODE194
• INLINECODE195: auto (default), separator (split on ---)
• INLINECODE199: Output directory (default: same as input file)
• INLINECODE200: Card width in px (default: 1080)
• INLINECODE201: Card height in px (default: 1440)
• INLINECODE202: Device pixel ratio (default: 2)

Requires: puppeteer-core and marked (npm install -g puppeteer-core marked). Does NOT require XHS cookies — purely offline rendering.

Override Chrome path: Set CHROME_PATH environment variable if Chrome is not in the standard location.

redbook whoami

Check connection status. Verifies cookies are valid and shows the logged-in user.

CODEBLOCK29

redbook post (Limited)

Publish an image note. Frequently triggers captcha (type=124) on the creator API. Image upload works, but the publish step is unreliable. For posting, consider using browser automation instead.

CODEBLOCK30

Options:

• - --title <title>: Note title (required)
• INLINECODE210: Note body text (required)
• INLINECODE211: Image file paths (required, at least one)
• INLINECODE212: Search and attach a topic hashtag
• INLINECODE213: Publish as private note

Global Options

All commands accept:

• - --cookie-source <browser>: chrome (default), safari, INLINECODE217
• INLINECODE218: Chrome profile directory name (e.g., "Profile 1"). Auto-discovered if omitted.
• INLINECODE219: Output as JSON

---

Technical Reference

xsec_token — Required for Reading & Sharing Notes

The XHS API requires a valid xsec_token to fetch note content. Without it, read, comments, and analyze-viral return {}.

The same token is also required for shareable URLs. Any https://www.xiaohongshu.com/explore/<id> URL without ?xsec_token=...&xsec_source=... is 302-redirected by XHS's anti-scrape layer to https://www.xiaohongshu.com/404/sec_*?source=xhs_sec_server&originalUrl=.... This affects anyone who clicks the URL — Safari, iOS link previews, agent action buttons, etc.

webUrl — use this (since v0.7.0):

Every note-returning command (feed, search, user-posts, favorites, board, read, post) now includes a webUrl field with the token baked in and the correct xsec_source. Consumers should use webUrl directly — do not construct URLs by hand.

CODEBLOCK31

INLINECODE239 is set per-command: pc_feed, pc_search, pc_user, pc_board, pc_share.

Key rules:

1. 1. Tokens expire. A URL with ?xsec_token=... from a previous session will return {}. Never cache or reuse old URLs.
2. search and feed always return fresh tokens. Every item includes a valid xsec_token + a pre-built webUrl.
3. noteId alone returns {}. Running redbook read <noteId> without a token almost always fails.

The correct workflow — always search first:

CODEBLOCK32

For agents: Prefer webUrl from the response. When only a bare noteId is available, search first to obtain a fresh token, then use the returned webUrl.

Commands that need xsec_token: read, comments, analyze-viral
Commands that do NOT need xsec_token: search, user, user-posts, feed, whoami, INLINECODE263

Chinese Number Formats in API Responses

The XHS API returns abbreviated numbers with Chinese unit suffixes:

API value	Actual number
INLINECODE264	15,000
INLINECODE265

24,000 |
| "1.2亿" | 120,000,000 |
| "115" | 115 |

INLINECODE268 = ×10,000. 亿 = ×100,000,000. Numbers under 10,000 are plain integers as strings.

The analyze-viral command handles this automatically. When parsing --json output manually, watch for these suffixes in interact_info fields (liked_count, collected_count, etc.).

Error Handling

Error	Meaning	Fix
INLINECODE275 empty response	Missing or expired xsec_token	Search first to get a fresh token
"No 'a1' cookie"

Not logged into XHS in browser | Log into xiaohongshu.com in Chrome | | "Session expired" | Cookie too old | Re-login in Chrome | | "NeedVerify" / captcha | Anti-bot triggered | Wait and retry, or reduce request frequency | | "IP blocked" (300012) | Rate limited | Wait or switch network |

---

Output Format Guidance

When producing analysis reports, use these formats:

Data tables: Markdown tables with exact field mappings. Always include the metric unit.

Heatmaps: ASCII bar charts for cross-topic comparison:
CODEBLOCK33

Creator comparison: Structured table with both quantitative metrics and qualitative style assessment.

Final reports: Use this section order:

1. 1. Market Overview (demand signals, content velocity)
2. Keyword Landscape (engagement matrix, opportunity tiers)
3. Cross-Topic Heatmap (topic × scene intersections)
4. Audience Persona (demographics, intent, preferences)
5. Competitive Landscape (creator profiles, strategy patterns)
6. Content Opportunities (tiered recommendations with data backing)
7. Content Ideas (specific hooks, angles, targets)

Programmatic API

CODEBLOCK34

Requirements

• - Node.js >= 22
• Logged into xiaohongshu.com in Chrome (or Safari/Firefox with --cookie-source)
• macOS (cookie extraction uses native keychain access)
• For card rendering only: puppeteer-core and marked (npm install -g puppeteer-core marked). Uses your existing Chrome — no additional browser download.