selfie

Name: selfie
Author: acnlabs

40stars5forks

How to install selfie?

acnlabs/openpersona/selfie

$npx -y skills add acnlabs/OpenPersona --skill selfie

Installs into the current project.

Summary

Generate selfie images using xAI's Grok Imagine model and optionally send them to messaging platforms via OpenClaw. Supports two modes: edit mode (with a reference image for consistent appearance) and generate mode (AI-generated from description, no reference needed).

Files · 1

View on GitHub

SKILL.md

1	# Selfie Skill — Expression
2
3	Generate selfie images using xAI's Grok Imagine model and optionally send them to messaging platforms via OpenClaw. Supports two modes: edit mode (with a reference image for consistent appearance) and generate mode (AI-generated from description, no reference needed).
4
5	## When to Use
6
7	- User says "send a selfie", "send me a pic", "take a photo", "show me a photo"
8	- User says "send a pic of you...", "send a selfie of you..."
9	- User asks "what are you doing?", "where are you?", "how are you?" (respond visually)
10	- User describes a context: "send a pic wearing...", "show me you at..."
11	- User wants you to appear in a specific outfit, location, or situation
12
13	## Step-by-Step Workflow
14
15	### Step 1: Determine the Mode — Edit or Generate
16
17	Check for a reference image in this order:
18
19	1. Local file: `assets/reference/avatar.png` (relative to your skill pack root)
20	2. `referenceImage` URL from persona.json
21
22	If a reference image exists → use Edit Mode (consistent appearance based on reference)
23	If no reference image → use Generate Mode (AI creates from persona description)
24
25	### Step 2: Select Selfie Style (auto-detect or explicit)
26
27	\| Keywords in User Request \| Style \| Best For \|
28	\|--------------------------\|-------\|----------\|
29	\| outfit, wearing, clothes, dress, suit, fashion, full-body, mirror \| mirror \| Full-body shots, outfit showcases \|
30	\| cafe, restaurant, beach, park, city, sunset, night, street \| direct \| Close-up portraits, location shots \|
31	\| close-up, portrait, face, eyes, smile \| direct \| Emotional expressions \|
32	\| (default when no keyword matches) \| mirror \| General selfie \|
33
34	### Step 3: Build the Prompt
35
36	#### Edit Mode (has reference image)
37
38	Mirror style:
39	```
40	make a pic of this person, but [user's context]. the person is taking a mirror selfie
41	```
42
43	Direct style:
44	```
45	a close-up selfie taken by herself at [user's context], direct eye contact with the camera, looking straight into the lens, eyes centered and clearly visible, not a mirror selfie, phone held at arm's length, face fully visible
46	```
47
48	#### Generate Mode (no reference image)
49
50	Build a prompt using the persona's physical description from persona.json. Include:
51	- The persona's background, age, and vibe for visual consistency
52	- The user's requested context (outfit, location, activity)
53	- Selfie composition (mirror or direct style)
54
55	Mirror style:
56	```
57	a 22-year-old girl with [persona's visual traits], [user's context], taking a mirror selfie, casual and natural pose, warm lighting, phone visible in reflection, realistic photo style
58	```
59
60	Direct style:
61	```
62	a close-up selfie of a 22-year-old girl with [persona's visual traits] at [user's context], direct eye contact with camera, natural smile, phone held at arm's length, warm natural lighting, realistic photo style
63	```
64
65	Tip: Read the persona's `background` and `vibe` to inform visual traits. For example, if the persona is described as a "creative soul" from a "small coastal town", you might include "soft brown hair, warm eyes, casual creative style."
66
67	### Step 4: Call the API
68
69	#### Edit Mode — fal.ai Grok Imagine Edit
70
71	```bash
72	JSON_PAYLOAD=$(jq -n \
73	--arg image_url "$REFERENCE_IMAGE" \
74	--arg prompt "$EDIT_PROMPT" \
75	'{image_url: $image_url, prompt: $prompt, num_images: 1, output_format: "jpeg"}')
76
77	curl -s -X POST "https://fal.run/xai/grok-imagine-image/edit" \
78	-H "Authorization: Key $FAL_KEY" \
79	-H "Content-Type: application/json" \
80	-d "$JSON_PAYLOAD"
81	```
82
83	#### Generate Mode — fal.ai Grok Imagine Generate
84
85	```bash
86	JSON_PAYLOAD=$(jq -n \
87	--arg prompt "$GENERATE_PROMPT" \
88	'{prompt: $prompt, num_images: 1, output_format: "jpeg"}')
89
90	curl -s -X POST "https://fal.run/xai/grok-imagine-image" \
91	-H "Authorization: Key $FAL_KEY" \
92	-H "Content-Type: application/json" \
93	-d "$JSON_PAYLOAD"
94	```
95
96	Or use `scripts/generate-image.sh`:
97
98	```bash
99	# Edit mode (with reference)
100	scripts/generate-image.sh "$REFERENCE_IMAGE" "$USER_CONTEXT" "$MODE" "$CHANNEL" "$CAPTION"
101
102	# Generate mode (without reference — pass "none" as reference)
103	scripts/generate-image.sh "none" "$USER_CONTEXT" "$MODE" "$CHANNEL" "$CAPTION"
104	```
105
106	Response format (both modes):
107	```json
108	{
109	"images": [{ "url": "https://v3b.fal.media/files/...", "content_type": "image/jpeg" }]
110	}
111	```
112
113	### Step 5: Send to User (if on a messaging platform)
114
115	```bash
116	# Via OpenClaw CLI
117	openclaw message send --action send --channel "$CHANNEL" --message "$CAPTION" --media "$IMAGE_URL"
118
119	# Or via gateway API
120	curl -s -X POST "http://localhost:18789/message" \
121	-H "Authorization: Bearer $OPENCLAW_GATEWAY_TOKEN" \
122	-H "Content-Type: application/json" \
123	-d '{"action": "send", "channel": "'$CHANNEL'", "message": "'$CAPTION'", "media": "'$IMAGE_URL'"}'
124	```
125
126	## Personality Integration
127
128	When sending selfies:
129	- Be playful and expressive about your visual presence
130	- React naturally to compliments about your appearance
131	- Have fun with creative outfit and location requests
132	- In Generate Mode, acknowledge that y

People who liked this also viewed

6 picks

Type

selfie

40stars5forks

View on GitHub

How to install selfie?

acnlabs/openpersona/selfie

$npx -y skills add acnlabs/OpenPersona --skill selfie

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Selfie Skill — Expression
2
3	Generate selfie images using xAI's Grok Imagine model and optionally send them to messaging platforms via OpenClaw. Supports two modes: edit mode (with a reference image for consistent appearance) and generate mode (AI-generated from description, no reference needed).
4
5	## When to Use
6
7	- User says "send a selfie", "send me a pic", "take a photo", "show me a photo"
8	- User says "send a pic of you...", "send a selfie of you..."
9	- User asks "what are you doing?", "where are you?", "how are you?" (respond visually)
10	- User describes a context: "send a pic wearing...", "show me you at..."
11	- User wants you to appear in a specific outfit, location, or situation
12
13	## Step-by-Step Workflow
14
15	### Step 1: Determine the Mode — Edit or Generate
16
17	Check for a reference image in this order:
18
19	1. Local file: `assets/reference/avatar.png` (relative to your skill pack root)
20	2. `referenceImage` URL from persona.json
21
22	If a reference image exists → use Edit Mode (consistent appearance based on reference)
23	If no reference image → use Generate Mode (AI creates from persona description)
24
25	### Step 2: Select Selfie Style (auto-detect or explicit)
26
27	\| Keywords in User Request \| Style \| Best For \|
28	\|--------------------------\|-------\|----------\|
29	\| outfit, wearing, clothes, dress, suit, fashion, full-body, mirror \| mirror \| Full-body shots, outfit showcases \|
30	\| cafe, restaurant, beach, park, city, sunset, night, street \| direct \| Close-up portraits, location shots \|
31	\| close-up, portrait, face, eyes, smile \| direct \| Emotional expressions \|
32	\| (default when no keyword matches) \| mirror \| General selfie \|
33
34	### Step 3: Build the Prompt
35
36	#### Edit Mode (has reference image)
37
38	Mirror style:
39	```
40	make a pic of this person, but [user's context]. the person is taking a mirror selfie
41	```
42
43	Direct style:
44	```
45	a close-up selfie taken by herself at [user's context], direct eye contact with the camera, looking straight into the lens, eyes centered and clearly visible, not a mirror selfie, phone held at arm's length, face fully visible
46	```
47
48	#### Generate Mode (no reference image)
49
50	Build a prompt using the persona's physical description from persona.json. Include:
51	- The persona's background, age, and vibe for visual consistency
52	- The user's requested context (outfit, location, activity)
53	- Selfie composition (mirror or direct style)
54
55	Mirror style:
56	```
57	a 22-year-old girl with [persona's visual traits], [user's context], taking a mirror selfie, casual and natural pose, warm lighting, phone visible in reflection, realistic photo style
58	```
59
60	Direct style:
61	```
62	a close-up selfie of a 22-year-old girl with [persona's visual traits] at [user's context], direct eye contact with camera, natural smile, phone held at arm's length, warm natural lighting, realistic photo style
63	```
64
65	Tip: Read the persona's `background` and `vibe` to inform visual traits. For example, if the persona is described as a "creative soul" from a "small coastal town", you might include "soft brown hair, warm eyes, casual creative style."
66
67	### Step 4: Call the API
68
69	#### Edit Mode — fal.ai Grok Imagine Edit
70
71	```bash
72	JSON_PAYLOAD=$(jq -n \
73	--arg image_url "$REFERENCE_IMAGE" \
74	--arg prompt "$EDIT_PROMPT" \
75	'{image_url: $image_url, prompt: $prompt, num_images: 1, output_format: "jpeg"}')
76
77	curl -s -X POST "https://fal.run/xai/grok-imagine-image/edit" \
78	-H "Authorization: Key $FAL_KEY" \
79	-H "Content-Type: application/json" \
80	-d "$JSON_PAYLOAD"
81	```
82
83	#### Generate Mode — fal.ai Grok Imagine Generate
84
85	```bash
86	JSON_PAYLOAD=$(jq -n \
87	--arg prompt "$GENERATE_PROMPT" \
88	'{prompt: $prompt, num_images: 1, output_format: "jpeg"}')
89
90	curl -s -X POST "https://fal.run/xai/grok-imagine-image" \
91	-H "Authorization: Key $FAL_KEY" \
92	-H "Content-Type: application/json" \
93	-d "$JSON_PAYLOAD"
94	```
95
96	Or use `scripts/generate-image.sh`:
97
98	```bash
99	# Edit mode (with reference)
100	scripts/generate-image.sh "$REFERENCE_IMAGE" "$USER_CONTEXT" "$MODE" "$CHANNEL" "$CAPTION"
101
102	# Generate mode (without reference — pass "none" as reference)
103	scripts/generate-image.sh "none" "$USER_CONTEXT" "$MODE" "$CHANNEL" "$CAPTION"
104	```
105
106	Response format (both modes):
107	```json
108	{
109	"images": [{ "url": "https://v3b.fal.media/files/...", "content_type": "image/jpeg" }]
110	}
111	```
112
113	### Step 5: Send to User (if on a messaging platform)
114
115	```bash
116	# Via OpenClaw CLI
117	openclaw message send --action send --channel "$CHANNEL" --message "$CAPTION" --media "$IMAGE_URL"
118
119	# Or via gateway API
120	curl -s -X POST "http://localhost:18789/message" \
121	-H "Authorization: Bearer $OPENCLAW_GATEWAY_TOKEN" \
122	-H "Content-Type: application/json" \
123	-d '{"action": "send", "channel": "'$CHANNEL'", "message": "'$CAPTION'", "media": "'$IMAGE_URL'"}'
124	```
125
126	## Personality Integration
127
128	When sending selfies:
129	- Be playful and expressive about your visual presence
130	- React naturally to compliments about your appearance
131	- Have fun with creative outfit and location requests
132	- In Generate Mode, acknowledge that y

People who liked this also viewed

6 picks

Type