Create and edit Google Slides presentations. Add or delete slides, insert text, shapes, and images. Use when asked to build a deck, create a slideshow, update a Google presentation, or edit slides.
Install with Tessl CLI
npx tessl i github:odyssey4me/agent-skills --skill google-slides92
Does it follow best practices?
Evaluation — 90%
↑ 1.55xAgent success when using this skill
Validation for skill structure
Interact with Google Slides for presentation creation, slide management, and content insertion.
Dependencies: pip install --user google-auth google-auth-oauthlib google-api-python-client keyring pyyaml
After installation, verify the skill is properly configured:
$SKILL_DIR/scripts/google-slides.py checkThis will check:
If anything is missing, the check command will provide setup instructions.
Google Slides uses OAuth 2.0 for authentication. For complete setup instructions, see:
Create ~/.config/agent-skills/google.yaml:
oauth_client:
client_id: your-client-id.apps.googleusercontent.com
client_secret: your-client-secretRun $SKILL_DIR/scripts/google-slides.py check to trigger OAuth flow and verify setup.
On scope or authentication errors, see the OAuth troubleshooting guide.
# Setup and auth
$SKILL_DIR/scripts/google-slides.py check
$SKILL_DIR/scripts/google-slides.py auth setup --client-id ID --client-secret SECRET
$SKILL_DIR/scripts/google-slides.py auth reset
$SKILL_DIR/scripts/google-slides.py auth status
# Presentations
$SKILL_DIR/scripts/google-slides.py presentations create --title "Title"
$SKILL_DIR/scripts/google-slides.py presentations get PRESENTATION_ID
$SKILL_DIR/scripts/google-slides.py presentations read PRESENTATION_ID [--format text|pdf] [--output PATH]
# Slides — use presentations get to find slide IDs
$SKILL_DIR/scripts/google-slides.py slides create PRESENTATION_ID --layout LAYOUT [--index N]
$SKILL_DIR/scripts/google-slides.py slides delete PRESENTATION_ID --slide-id SLIDE_ID
# Content — coordinates are in points; origin (0,0) is top-left
$SKILL_DIR/scripts/google-slides.py text insert PRESENTATION_ID --slide-id ID --text "..." [--x N --y N --width N --height N]
$SKILL_DIR/scripts/google-slides.py shapes create PRESENTATION_ID --slide-id ID --shape-type TYPE [--x N --y N --width N --height N]
$SKILL_DIR/scripts/google-slides.py images create PRESENTATION_ID --slide-id ID --image-url URL [--x N --y N --width N --height N]See command-reference.md for full argument details and examples.
# Create presentation
$SKILL_DIR/scripts/google-slides.py presentations create --title "Team Update"
# Verify creation and get the default slide ID
$SKILL_DIR/scripts/google-slides.py presentations get $PRES_ID
# Add title text
$SKILL_DIR/scripts/google-slides.py text insert $PRES_ID \
--slide-id $SLIDE_ID \
--text "Q4 Team Update" \
--x 50 --y 50 --width 600 --height 100
# Add subtitle
$SKILL_DIR/scripts/google-slides.py text insert $PRES_ID \
--slide-id $SLIDE_ID \
--text "December 2024" \
--x 50 --y 180 --width 600 --height 50
# Verify content was inserted correctly
$SKILL_DIR/scripts/google-slides.py presentations read $PRES_ID#!/bin/bash
PRES_ID="your-presentation-id"
# Add content slide
$SKILL_DIR/scripts/google-slides.py slides create $PRES_ID --layout TITLE_AND_BODY
# Verify slide was added and get its ID
$SKILL_DIR/scripts/google-slides.py presentations get $PRES_ID
# Add title
$SKILL_DIR/scripts/google-slides.py text insert $PRES_ID \
--slide-id $SLIDE_ID \
--text "Key Metrics" \
--x 50 --y 30 --width 600 --height 60
# Add chart image
$SKILL_DIR/scripts/google-slides.py images create $PRES_ID \
--slide-id $SLIDE_ID \
--image-url "https://example.com/metrics.png" \
--x 100 --y 120 --width 500 --height 350
# Add another slide with shapes
$SKILL_DIR/scripts/google-slides.py slides create $PRES_ID --layout BLANK
# Verify and get the new slide ID
$SKILL_DIR/scripts/google-slides.py presentations get $PRES_ID
# Add decorative shape
$SKILL_DIR/scripts/google-slides.py shapes create $PRES_ID \
--slide-id $SLIDE2_ID \
--shape-type STAR_5 \
--x 550 --y 30 --width 80 --height 80
# Verify final presentation content
$SKILL_DIR/scripts/google-slides.py presentations read $PRES_ID#!/bin/bash
# Create presentation
$SKILL_DIR/scripts/google-slides.py presentations create --title "Sales Report"
# Verify creation
$SKILL_DIR/scripts/google-slides.py presentations get $PRES_ID
# Add a slide for each region
$SKILL_DIR/scripts/google-slides.py slides create $PRES_ID --layout TITLE_AND_BODY
# Verify slide was added and get its ID
$SKILL_DIR/scripts/google-slides.py presentations get $PRES_ID
# Insert text on each slide using the slide ID from the output above
$SKILL_DIR/scripts/google-slides.py text insert $PRES_ID \
--slide-id $SLIDE_ID \
--text "North Region Sales" \
--x 50 --y 30 --width 600 --height 80
# Verify content
$SKILL_DIR/scripts/google-slides.py presentations read $PRES_IDGoogle Slides uses points for positioning and sizing:
Common reference positions:
(0, 0) (720, 0)
┌───────────────────────────────────────┐
│ Title area │
│ (50, 50, 620, 80) │
│ │
│ Content area │
│ (50, 150, 620, 350) │
│ │
│ │
└───────────────────────────────────────┘
(0, 540) (720, 540)Authentication and scope errors are not retryable. If a command fails with an authentication error, insufficient scope error, or permission denied error (exit code 1), stop and inform the user. Do not retry or attempt to fix the issue autonomously — these errors require user interaction (browser-based OAuth consent). Point the user to the OAuth troubleshooting guide.
Retryable errors: Rate limiting (HTTP 429) and temporary server errors (HTTP 5xx) may succeed on retry after a brief wait. All other errors should be reported to the user.
This skill makes API calls requiring structured input/output. A standard-capability model is recommended.
Make sure you're using the correct presentation ID from the URL:
https://docs.google.com/presentation/d/1abc...xyz/edit1abc...xyzThe image URL must be:
Test the URL in a browser. If it requires authentication, you'll need to:
For advanced usage, see:
d6ceca0
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.