BeautyPlus Skills & API

> {baseDir} placeholder. Every CLI command in this file uses

> {baseDir}/scripts/.... The host (skill loader) substitutes {baseDir}

> with this skill's installation directory at load time. If your host does

> not expand placeholders, replace {baseDir} with the absolute path to

> this skill folder, or run from inside the skill folder using scripts/....

When to Use This Skill

Activate whenever the user requests modifications, enhancements, or artistic

changes to a photo / image (path, URL, or attachment).

Trigger keywords: "make me look better", "change my clothes", "fix this

blurry photo", "make me smile", "enhance my body", "change my hair color /

style", "add a filter", "cosplay", "outfit change", "upscale", "virtual

try-on".

Effect KEY catalog (--task values)

The CLI --task value is always one of the effect KEYs below. Per-key

algorithm params are returned inline by POST /skill/consume.json

(invoke_spec) — never hard-code AIGC paths. If the server returns

Unknown invoke preset, the key is not in the tenant's invoke map; do not

invent params, check config or admin.

All KEYs operate on still images (.jpg / .jpeg / .png / .webp). The current catalog is image-only; the async / video flow

is reserved — see docs/video-reserved.md.

1. 💪 Body reshape

When: user explicitly asks to enhance, enlarge, or reshape breast or buttocks.

Tier rule: append _strong / _medium / _weak to the base KEY.

Default tier: _medium (e.g. breast_natural_medium).

2. 💇‍♀️ Hair editing

When: new hair color or new hairstyle.

3. 👗 Outfit change (AI clothes)

When: user asks to change clothes, try on different styles, or dress up

for a specific occasion / cosplay.

Formal / Evening

Vacation / Casual

Cosplay / Fantasy

Party / Futuristic

Sports / Activewear

4. 🎭 Facial retouch & expression

When: retouch / apply a makeup style, or change the subject's expression.

5. 🖼️ Photo quality & art filters

When: image is blurry / noisy / old / low-resolution, or the user wants

a camera vibe / lighting filter.

Defaults & disambiguation (MANDATORY)

Use this table when the user's intent is broad or under-specified. Do not

guess a wrong KEY.

API submission path (MANDATORY)

• New jobs must go through

python3 {baseDir}/scripts/beautyplus_ai.py run-task … (or the same

run-task embedded inside spawn-run-task → sessions_spawn for the

reserved video flow).

• Do not hand-craft HTTP to the skill's wapi gateway or to

AIGC / invoke endpoints — that skips POST /skill/consume.json (quota and

permission) and breaks the supported pipeline. This skill does not emit

debug curl for API calls.

• query-task --task-id is for resuming polling on an existing

full task_id only; it does not replace run-task for a new

submission.

Billing and user-facing claims (MANDATORY)

• Each successful run-task (including inside a sessions_spawn worker)

triggers server-side quota / credit consumption for the BP_AK

tenant. This is a paid, metered commercial API.

• Do not state or imply that the service is free, costs nothing, has an

unlimited trial, or invent prices, plan names, promotions, or trial rules.

• Allowed wording: processing uses the BeautyPlus account quota tied to

the configured keys; billing and plans are per the user's console or

administrator.

• On quota / membership errors, surface the server detail and

pricing_url (see Step 3 error table). Never add "free" or zero-cost

wording, even on success.

Step 0 — Pre-flight check (MANDATORY)

Verify AK / SK are configured. Only run this command at this stage; do

not read other Python sources first.

python3 {baseDir}/scripts/beautyplus_ai.py preflight

• Output ok → continue to Step 1.
• Output missing → stop and prompt the user. Use the channel-appropriate

template in docs/credentials-prompt.md

(plain text by default; Feishu interactive card when the host exposes a

Feishu account).

Step 1 — Pick effect KEY and input

1. Map the user's intent to one effect KEY from the catalog above. Apply

the Defaults & disambiguation table when the request is broad.

1. Confirm the input file location. If the user already gave a path or URL,

continue to Step 2 without asking again.

1. Reply immediately to acknowledge, e.g. *"🖼️ Processing — please wait

a moment…"*

Getting media from IM messages

Full detail: docs/im-attachments.md.

python3 {baseDir}/scripts/beautyplus_ai.py resolve-input --file /tmp/saved.jpg --output-dir /tmp
# or: --url, --telegram-file-id, --feishu-image-key + --feishu-message-id

Use the JSON path field as --input.

--input as a URL: quote the full URL in the shell to avoid &

splitting. Defaults: 120 s read timeout, 100 MB max — override with

MT_AI_URL_READ_TIMEOUT, MT_AI_URL_CONNECT_TIMEOUT, MT_AI_URL_MAX_BYTES.

For flaky or large links, use resolve-input --url first and pass the local

path.

Step 2 — Install dependencies

python3 {baseDir}/scripts/beautyplus_ai.py install-deps

If dependencies are already installed this step is quick; then continue to

Step 3.

Step 3 — Run the task

All listed effect KEYs are image jobs — use the inline run-task flow

(§3a). Video keys are reserved; see docs/video-reserved.md.

3a — Inline run-task (blocking, default)

python3 {baseDir}/scripts/beautyplus_ai.py run-task \
  --task "<effect_key>" \
  --input "<image_url_or_path>"

Replace (e.g. hair_soft_waves) and

with real values. Default params include rsp_media_type: url. For custom

JSON params:

python3 {baseDir}/scripts/beautyplus_ai.py run-task \
  --task "<effect_key>" \
  --input "<url_or_path>" \
  --params '{"parameter":{"rsp_media_type":"url"}}'

When run-task exits 0

Stdout is JSON containing:

• skill_status: "completed" — job done.
• output_urls — ordered http(s) links (extracted from

data.result.urls, images, media_info_list, etc.).

• primary_result_url — same as output_urls[0]; convenient for delivery.
• task_id — full task id when the API returns one. Keep it for support /

recovery; do not truncate. Some sync completions may omit it.

• agent_instruction — short reminder for the model.
• meta / data — full API payload for debugging.

You must:

1. Send the user a short natural-language summary of what was done.
2. Complete Step 4 delivery using primary_result_url or output_urls[0],

unless the user explicitly asked for the URL only.

Never end the turn with raw JSON alone. Never resubmit run-task for the

same task_id — use query-task to resume polling.

When run-task exits non-zero

Stdout has skill_status: "failed" with an error field. Map it as:

Never dump raw JSON to the user. Never retry run-task for credit /

membership errors.

3b — Resume polling (query-task)

When you already have a full task_id (from previous stdout JSON, polling

checkpoint, or task_id=... lines in stderr) and the job may still be

running, do not run run-task again — resume polling only:

python3 {baseDir}/scripts/beautyplus_ai.py query-task \
  --task-id "<full_task_id>"

Optional --task sets the task_name field in the success JSON for your

logs (default labels as query_task). Uses the same BP_AK / BP_SK and

remote config as the original submit. Stdout JSON and exit codes match

run-task.

3c — Last task and history

Local state lives under ~/.openclaw/workspace/beautyplus-ai/

(last_task.json, history/task_*.json, last 50 records). For async jobs,

last_task.json may briefly show skill_status: "polling" with a

task_id while the client is still polling — a checkpoint that lets

query-task resume if the process is killed mid-poll.

python3 {baseDir}/scripts/beautyplus_ai.py last-task
python3 {baseDir}/scripts/beautyplus_ai.py history

Use when the user asks whether a recent job finished, or wants a short

history summary. Do not expose raw secrets.

Step 4 — Deliver result to the channel

Required after skill_status: "completed". The CLI does not post to IM by

itself — send the processed image back on the user's platform, in the same

turn as the success summary.

Resolve deliver-to

Feishu

python3 {baseDir}/scripts/feishu_send_image.py \
  --image "<result_url>" \
  --to "<oc_xxx or ou_xxx>"

Telegram

TELEGRAM_BOT_TOKEN="$TELEGRAM_BOT_TOKEN" python3 {baseDir}/scripts/telegram_send_image.py \
  --image "<result_url>" \
  --to "<chat_id>" \
  --caption "✅ Done"

Discord

Download the result, then send with the message tool:

curl -L "<result_url>" -o /tmp/result_image.jpg

message(action="send", channel="discord", target="<channel_id>", filePath="/tmp/result_image.jpg")

For files over ~25 MB, send the result URL as a link instead.

WhatsApp / Signal / others

Use the message tool with media, or send the result URL directly.

Video delivery (reserved): see docs/video-reserved.md.

Multi-stage pipelines (chaining tasks)

When the user asks for more than one BeautyPlus step on the same media

(e.g. photo_restoration_v3 → dress_yellow_gown), treat each step as a

separate job with its own --task:

1. After stage A completes, pass primary_result_url (or output_urls[0])

as --input for the next stage. That is a new job, not a retry.

1. "Do not re-run run-task" means: do not resubmit the same

task_id. It does not forbid the next stage with a different effect KEY.

1. Deliver only after the final stage completes.

Video chains are reserved — see docs/video-reserved.md.

Notes

• Single business entrypoint: all algorithm runs and config fetches go

through beautyplus_ai.py; agents do not need to open client.py or

ai/api.py. See API submission path above for the no-direct-HTTP

rule.

• AK / SK loading: environment variables BP_AK / BP_SK first; if

unset, scripts/.env is read automatically.

• Client init pulls the latest algorithm config from the server; no

manual INVOKE setup is required.

• Bot token safety: pass TELEGRAM_BOT_TOKEN, DISCORD_BOT_TOKEN, etc.

only via environment variables — never as CLI arguments.

• URL input errors may report HTTP 403 (expired signed URL) or timeout —

see MT_AI_URL_* env vars in Step 1.

Further reading

• BeautyPlus Developers — API and skill key integration, use cases.
• BeautyPlus — product overview.
• README.md
• docs/video-reserved.md — async / video flow (reserved).
• docs/credentials-prompt.md — preflight: missing user prompt templates.
• docs/multi-platform.md — delivery details (Feishu, Telegram, Discord, …).
• docs/im-attachments.md — IM attachments and resolve-input.
• docs/errors-and-polling.md — polling, timeouts, failure codes.