Agent skill

Pixiv CLI

by FlanChanXwO in FlanChanXwO/pixiv-cli

Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized…

MITAuto-check passedBackend & APIs

Install Pixiv CLI

skills CLI
$ npx skills add FlanChanXwO/pixiv-cli --skill pixiv-cli -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install FlanChanXwO/pixiv-cli pixiv-cli --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/FlanChanXwO/pixiv-cli.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/pixiv-cli .claude/skills/pixiv-cli && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
pixiv-cli
GitHub stars
134
Token cost
~2.8k tokens
SKILL.md length
1,316 words
Files
7 (incl. references)
Skills in repo
11
Repo updated
First seen
Licence
MIT

At a glance

Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized…

  • Generic illustration
  • SKILL.md covers Preflight and routing, Secrets and side effects, Output and complete results and Authentication and service…, plus 3 more sections
  • Needs PIXIV_REFRESH_TOKEN and SAUCENAO_API_KEY
  • Download requests

What it does

Pixiv CLI is an agent skill from FlanChanXwO/pixiv-cli. Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized bookmark, follow, comment, account, or download actions. Use only for an explicit Pixiv or pixiv-cli request, a pixiv command, or a Pixiv URL/ID with clear operation context. Do not trigger for generic illustration, artist, image-search, or download requests. Verify syntax with pixiv COMMAND --help.

Its SKILL.md is about 2.8k tokens, which your agent loads only when the skill is triggered. The skill folder holds 8 other files, including reference files (for example `agents/openai.yaml`, `references/auth.md` and `references/discover.md`).

It sits in Backend & APIs. The repository describes itself as: Pixiv, in your terminal — a CLI, MCP server, and Go SDK for discovery, accounts, creators, collections, and downloads. The licence is MIT.

When your agent uses it

  • Generic illustration
  • Download requests

Example prompts

  • “/pixiv-cli”

Requirements

  • A credential in PIXIV_REFRESH_TOKEN
  • A credential in SAUCENAO_API_KEY

What it can do on your machine

Read from SKILL.md and the folder at commit 2b3ccbb. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • PIXIV_REFRESH_TOKEN
    • SAUCENAO_API_KEY

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Pixiv CLI loads about 2.8k tokens when it runs, and up to ~12k if it reads all its reference files. Until then it costs about 126 tokens; SKILL.md has 1,316 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~126
When it runs · the whole SKILL.md, loaded when a task matches
~2.8k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~12k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from FlanChanXwO/pixiv-cli at commit 2b3ccbb, republished under its MIT licence (© FlanChanXwO). 1,316 words, ~2,767 tokens.

Download SKILL.mdSave it as .claude/skills/pixiv-cli/SKILL.md (or your agent's skills folder). This skill also uses 6 other files; get the full folder from GitHub.
name
pixiv-cli
description
Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized bookmark, follow, comment, account, or download actions. Use only for an explicit Pixiv or pixiv-cli request, a pixiv command, or a Pixiv URL/ID with clear operation context. Do not trigger for generic illustration, artist, image-search, or download requests. Verify syntax with pixiv COMMAND --help.
slug
pixiv-cli
version
1.1.1
displayName
Pixiv CLI
summary
Safely operate Pixiv with the pixiv-cli binary for discovery, account actions, and downloads.
license
MIT
homepage
https://github.com/FlanChanXwO/pixiv-cli
tags
pixiv, cli, mcp

Pixiv CLI Operator

Use the installed pixiv binary. This bundle describes operation, privacy, and semantic traps, not repository development. It works without a source checkout or personal maintenance skills.

Preflight and routing

Run pixiv --version and require a successful pixiv VERSION line. A missing or non-executable binary is a blocker, not authorization to install it. Inspect the relevant command's --help before execution and read only the needed reference:

TaskReference
Explicit installation or repairinstall.md
Login, import/export, backup/restore, account poolauth.md
Search, reverse image, detail, users, feeds, rankings, bookmarksdiscover.md
Single, multi-page, batch, or animated downloadsdownload.md
Authentication, proxy, output, or incomplete resultstroubleshooting.md

Do not enumerate accounts on every turn. Use pixiv auth list --json only when an account/authentication decision requires it; both accounts: null and accounts: [] mean none. Stored credentials do not prove current validity. Run networked auth check only when needed; auth refresh rotates credentials and refreshes profile/Premium metadata, so requires an explicit maintenance request.

Secrets and side effects

Refresh tokens and export bundles are plaintext secrets. Keep them out of commentary, results, logs, and diagnostics. Before any import/export or browser login, follow auth.md: an explicitly requested bare stdout export has a narrow disclosure/confirmation workflow; every other export uses a private output or direct consumer pipeline. Do not inspect credential-store contents to debug an account.

Do not ask for an undisclosed token in chat or launch a hidden prompt in an agent terminal the user cannot type into. If the user already disclosed a token and explicitly requests positional import, first explain the additional tool/process-record exposure, safely quote it, and never repeat it in results. Non-TTY import reads stdin automatically; there is no --stdin flag. Strict bundle decoding must not fall back to OAuth.

OperationRequired scope
Discovery and read commandsExecute only as needed for the task; do not add account probes or extra result traversal
auth list/checkOnly an account decision or needed validation
auth refreshExplicit account-maintenance request
Bookmark/follow/comment mutationsExplicit current targets, types, and action; state the scope before execution
downloadConfirm exact targets and destination for each invocation; a user URL expands the whole supported visual listing
auth use/remove, pool enable/disable, config set/unset, actual updateExplicit current state-change authorization; it does not carry to later targets
auth loginExplicit request and a user present for browser OAuth; follow the documented desktop/remote handoff
mcp or fanbox mcpExplicit server-start request; these are long-lived stdio servers, not preflight probes

config path can create baseline configuration if missing. A successful read-looking command is not necessarily filesystem-side-effect-free. update --check --json checks only; actual installation does not emit JSON and requires authorization.

Output and complete results

Use the requested scope and command-specific flags. A supported positive --limit N requests N logical results; reaching N is not proof that only N exist. Use --limit 0 only for an explicit exhaustive request. --page requires positive --limit; an omitted limit normally means one logical batch, not the entire corpus. Local filters can skip empty upstream batches; do not invent request caps.

Use text for a small display-only result. Visual listing commands emit canonical NDJSON automatically when stdout is piped; --ndjson makes that explicit. --json is one complete result document where supported and cannot combine with --ndjson. Check status before parsing; failed commands may produce ordinary stderr and empty stdout. Do not install processing tools merely to format output.

text
pixiv search "landscape" --type artwork --limit 20 --ndjson | pixiv detail
pixiv search ./image.png --provider all --ndjson | pixiv detail --ndjson
pixiv detail ARTWORK_ID --json
pixiv search "fantasy" --type novel --limit 10 --json
pixiv user search "NAME" --limit 10 --json
pixiv ranking --type novel --mode week --limit 10
pixiv recommended --type artwork --limit 10

Use these examples only for the requested operation and inspect selected records before any downstream action. Aggregate search --json is not a record stream. In detail record mode, record types choose artwork/novel/user endpoints; explicit --type is only a compatibility constraint. Record --json produces an array, not a single-object promise.

Public positional commands can fill one missing value from non-TTY stdin. Only a final newline is removed; an explicit positional value wins, and two missing required values do not trigger generic stdin reading. For detail, download, and bookmark/follow actions, an initial non-whitespace { selects strict record input; failures do not switch to raw ID input. - is ordinary text, not a universal stdin sentinel.

Authentication and service boundaries

Pixiv content is App-only. Use the selected local auth use account or an eligible database-managed account when the pool is enabled. There is no anonymous Web fallback and no per-command --uid/--refresh-token selection or PIXIV_REFRESH_TOKEN credential shortcut. The public SDK and MCP have their own explicit runtime credential contracts.

Pool enabled/strategy settings are configuration; schedulable accounts are managed by auth pool status|enable|disable. Reads/downloads may use the pool, but mutations must not be silently assigned to a different account. Removed settings fail explicitly; use the authorized cleanup described in auth.md, not an automatic account migration.

FANBOX is separate: use pixiv fanbox --help and its read-only creator/post/tag/feed/resource capabilities. Importing FANBOXSESSID with pixiv fanbox auth import or selecting an account with pixiv fanbox auth use --auto/an explicit UID is a separately authorized credential/state action, not automatic preflight. FANBOX MCP does not reuse the Pixiv pool. Its optional challenge-only solver does not carry ordinary API/resource traffic or receive the session; never send the session to a non-first-party host.

Show full SKILL.md (480 more words)Show less

Search and identity traps

  • Artwork --rating sfw|r18|r18g|mature|all filters normalized x_restrict locally and binds the filter to its cursor. It is not an upstream request field and is not supported for novel search. Artwork subtype, AI, resolution, aspect, and exact drawing-tool filters have their own validated scope.
  • Bookmark counts are not likes. auto/local filter candidates locally, best_effort is explicitly partial, and server fails without verified upstream support. Premium is not a local hard gate. Do not present partial candidates as a complete site-wide result.
  • recommended needs an entity kind and authentication; local artwork subtype filtering is not an upstream query. Rankings and feeds have command-specific kinds/modes; do not substitute a failed extended ranking with day.
  • User detail/list routes take numeric user IDs where documented; not every command accepts a user URL. Follow writes accept a positive numeric ID or canonical user record, and --restrict is public|private with public default. Use only verified IDs and supported URL namespaces.
  • detail --type novel --content is a compatibility flag whose unavailable App content endpoint returns content_unavailable; it is not permission to scrape a WebView. Plain novel detail remains metadata.

Comment operations

Comment reads preserve optional total and access_control, including numeric upstream comment_access_control; do not invent boolean permission fields. comment create|reply|stamp|delete requires an explicit artwork/novel type and positive numeric IDs, not URLs or all. Create/reply needs non-empty --comment; reply also needs --parent-comment-id, and stamp needs --stamp-id with optional text. Confirm the user's exact content and target before writing.

Create/reply/stamp return the upstream comment ID. Delete returns a success status without an automatic read-back. comment stamps is a non-paginated read with safe opaque stamp references. Bookmark/follow actions retain empty-success stdout; do not invent a JSON report.

Reverse search, proxies, and delivery

Reverse-search image sources are existing regular files or explicit HTTP(S) URLs; an invalid explicit URL is an error, never a keyword fallback. Binary image bytes on stdin do not select image mode. Provider/output/proxy flags apply; keyword filters, entity type, and pagination do not.

Providers are SauceNAO, ascii2d color/BoVW, and all. reverse_search_pixiv_only controls canonical Pixiv identities. Configure saucenao_api_key only through non-TTY stdin or SAUCENAO_API_KEY; reads redact it. Upload only authorized images and preserve provider privacy boundaries. One successful provider plus another failure can be a successful partial result with a warning; single/all-provider failure is nonzero. Generic reverse-search type=artwork records do not prove a Pixiv subtype.

--proxy and --no-proxy are mutually exclusive invocation overrides. Pixiv, FANBOX, reverse-search transport, and solver-browser settings have distinct scopes; the browser's proxy is not automatically inherited. Do not assume a private maintainer's loopback proxy exists. Read troubleshooting.md before changing persistent settings or solver configuration.

Downloads write files and have empty successful stdout; inspect exit status and the requested destination, following download.md. A legitimate long encode/download is not a timeout failure. Share generated files only through an available attachment API; otherwise report the source artwork URL and do not claim an image was sent.

© FlanChanXwO, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 6 other files (references) in skills/pixiv-cli of FlanChanXwO/pixiv-cli.

  • SKILL.md
  • agents/openai.yaml
  • references/auth.md
  • references/discover.md
  • references/download.md
  • references/install.md
  • references/troubleshooting.md

Open the folder on GitHubat commit 2b3ccbb

Compare with similar skills

Pixiv CLI next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Pixiv CLI compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Pixiv CLI this skillFlanChanXwO/pixiv-cli134—~2.8kAutomated safety check: PassMIT
Configuring Horizoncoollabsio/coolify63k4 repos~898Automated safety check: PassMIT
Nestjs Best Practicesrolling-scopes/rsschool-app10k6 repos~1.2kAutomated safety check: PassMIT
Sub2API AdminWei-Shaw/sub2api43k1 repos~717Automated safety check: PassLGPL-3.0
Firecrawl Build Onboardingfirecrawl/firecrawl190k1 repos~1.4kAutomated safety check: NotesISC
Obsidian BasesAtmosphere/atmosphere3.8k22 repos~3.2kAutomated safety check: PassApache-2.0

Similar skills

  • Configuring Horizon

    coollabsio/coolify

    A skill your agent uses whenever the user mentions Horizon by name in a Laravel context.

    63k GitHub starsUsed in 4 repos~898 tokens
    Backend & APIsAuto-check passed
  • Nestjs Best Practices

    rolling-scopes/rsschool-app

    NestJS best practices and architecture patterns for building production-ready applications.

    10k GitHub starsUsed in 6 repos~1.2k tokens
    Backend & APIsAuto-check passed
  • Sub2API Admin

    Wei-Shaw/sub2api

    Manages a Sub2API deployment from the command line: accounts, redeem and invitation codes, groups, proxies, imports, exports and raw admin API calls.

    43k GitHub starsUsed in 1 repo~717 tokens
    Backend & APIsAuto-check passed
  • Firecrawl Build Onboarding

    firecrawl/firecrawl

    Gets Firecrawl working in a project: signs you in through the browser, saves FIRECRAWL_API_KEY to .env and picks the first SDK or REST path.

    190k GitHub starsUsed in 1 repo~1.4k tokens
    Backend & APIsAuto-check: notes
  • Obsidian Bases

    Atmosphere/atmosphere

    Create and edit Obsidian Bases (.base files) with views, filters, formulas, and summaries.

    3.8k GitHub starsUsed in 22 repos~3.2k tokens
    Backend & APIsAuto-check passed
  • Fortify Development

    coollabsio/coolify

    ACTIVATE when the user works on authentication in Laravel. An agent skill from coollabsio/coolify.

    63k GitHub starsUsed in 4 repos~1.9k tokens
    Backend & APIsAuto-check passed

More from FlanChanXwO/pixiv-cli

All 11 skills in this repo
  • Pixiv CLI CI

    FlanChanXwO/pixiv-cli

    Diagnose pixiv-cli GitHub Actions failures and verify current-head quality, platform, PR-verification, native/browser, and release evidence.

    134 GitHub stars~1.3k tokensUpdated 2 days ago
    Auto-check passed
  • Pixiv CLI Code Commenting

    FlanChanXwO/pixiv-cli

    In pixiv-cli, write and review intent-focused code comments, API documentation, docstrings, and numbered workflow stages.

    134 GitHub stars~2.4k tokensUpdated 2 days ago
    Auto-check passed
  • Pixiv CLI Native

    FlanChanXwO/pixiv-cli

    Change or validate pixiv-cli Rust ugoira, cgo/FFI, committed static libraries, vendored dependencies, Linux ABI, and native runner evidence.

    134 GitHub stars~948 tokensUpdated 2 days ago
    Auto-check passed
  • Pixiv CLI PR

    FlanChanXwO/pixiv-cli

    Prepare, update, or verify a pixiv-cli pull request using its current template, trusted verification policy, reviewed diff, and head-specific check results.

    134 GitHub stars~802 tokensUpdated 2 days ago
    Auto-check passed
  • Pixiv CLI Docs

    FlanChanXwO/pixiv-cli

    Edit pixiv-cli documentation, AGENTS.md, client bridges, repository maintenance skills, or the distributed product skill.

    134 GitHub stars~997 tokensUpdated 2 days ago
    Auto-check passed
  • Pixiv CLI MCP Tool

    FlanChanXwO/pixiv-cli

    Add, change, or review a pixiv-cli Pixiv or FANBOX MCP tool, including schema, SDK routing, structured output, errors, and stdio behavior.

    134 GitHub stars~750 tokensUpdated 2 days ago
    Auto-check passed

Categories

Questions about Pixiv CLI

What does Pixiv CLI do?

Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized…. Pixiv CLI is an agent skill from FlanChanXwO/pixiv-cli. Operate the installed pixiv CLI to search works and users, reverse-search images with SauceNAO or ascii2d, inspect Pixiv references, read feeds and rankings, and perform explicitly authorized bookmark, follow, comment, account, or download actions.

When should I use Pixiv CLI?

Pixiv CLI fits situations like: generic illustration; download requests.

How do I install Pixiv CLI in Claude Code?

Run `npx skills add FlanChanXwO/pixiv-cli --skill pixiv-cli -a claude-code`. Or copy the skill folder (skills/pixiv-cli in FlanChanXwO/pixiv-cli) into .claude/skills/pixiv-cli in your project. Claude Code loads it when a task matches its description.

How do I install Pixiv CLI in Codex?

Run `npx skills add FlanChanXwO/pixiv-cli --skill pixiv-cli -a codex`. Or copy the skill folder (skills/pixiv-cli in FlanChanXwO/pixiv-cli) into .agents/skills/pixiv-cli in your project. Codex loads it when a task matches its description.

Can I use Pixiv CLI in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add FlanChanXwO/pixiv-cli --skill pixiv-cli -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/pixiv-cli, .gemini/skills/pixiv-cli, .github/skills/pixiv-cli and .opencode/skills/pixiv-cli in your project.

What does Pixiv CLI need to run?

Going by SKILL.md and its folder, Pixiv CLI needs credentials named PIXIV_REFRESH_TOKEN and SAUCENAO_API_KEY. Our summary lists: A credential in PIXIV_REFRESH_TOKEN; A credential in SAUCENAO_API_KEY.

Does Pixiv CLI access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Pixiv CLI safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Pixiv CLI use?

Pixiv CLI is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Pixiv CLI use?

About 2.8k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 9.4k tokens, read only when the agent opens those files.

What are the alternatives to Pixiv CLI?

Skills that share tags, products or a category with Pixiv CLI: Configuring Horizon (coollabsio/coolify, 63k stars), Nestjs Best Practices (rolling-scopes/rsschool-app, 10k stars), Sub2API Admin (Wei-Shaw/sub2api, 43k stars) and Firecrawl Build Onboarding (firecrawl/firecrawl, 190k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Pixiv CLI?

FlanChanXwO (a GitHub user) maintains it in FlanChanXwO/pixiv-cli, which has 134 GitHub stars. The repository holds 11 skills in this directory. The repository was last updated on October 6, 2026.

Source: FlanChanXwO/pixiv-cli on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.