Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/sacredvoid/skillkit/image-fetchernpx skills add sacredvoid/skillkit --skill image-fetchergit clone --depth 1 https://github.com/sacredvoid/skillkitWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00057 | $0.02297 |
| Opus 5 | $0.00028 | $0.01149 |
| Sonnet 5 | $0.00011 | $0.00459 |
| Haiku 4.5 | $0.00006 | $0.00230 |
Grade A, and why
image-fetcher scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s "https://api.unsplash.com/search/photos?query=QUERY&per_page=COUNT&orientation=landscape" \ Copies of this mod
1 near-identical copy found in the catalogue:
- image-fetcher — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 258 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Image Fetcher
Fetch relevant, free-to-use images from the web for any project context.
Config
Read config from ~/.claude/skills/image-fetcher/config.json:
{
"unsplash_key": "",
"pexels_key": "",
"pixabay_key": "",
"default_count": 5,
"output_dir": "assets/images"
}
All fields are optional. The skill works with zero configuration via the WebSearch fallback.
Load config silently at the start of every invocation:
cat ~/.claude/skills/image-fetcher/config.json 2>/dev/null || echo '{}'
Phase 1: Determine What to Search For
If the user provided input (argument after the skill name), use that directly as the search context. Skip to Phase 2.
If no input was provided, scan the current directory for context signals:
| Source | What to extract |
|---|---|
README.md / README |
Project name, description, purpose |
package.json / Cargo.toml / pyproject.toml |
Name, description, keywords |
CLAUDE.md |
Project conventions, domain |
| HTML/CSS files (first 3) | Page titles, headings, color themes |
Existing images in assets/, public/, static/, images/ |
What already exists (to avoid duplicates) |
Use Glob and Read to scan. Build a brief internal context summary:
CONTEXT:
project: [name or "unknown"]
domain: [web app / game / docs / CLI / etc.]
keywords: [extracted keywords]
existing_images: [list of what's already there]
If the directory scan yields nothing useful (empty dir, no descriptive files), use AskUserQuestion:
"What kind of images are you looking for?"
Options:
- Hero/banner images
- Product photos
- Abstract/background textures
- Icons/illustrations
- Other (describe)
Then ask a follow-up for specifics: subject, mood, color preference.
From the context, generate 1-3 concise search queries (e.g., "minimalist coffee shop interior", "abstract tech gradient background").
Present the queries to the user:
Based on your project context, I'll search for:
- "minimalist coffee shop interior"
- "coffee beans close-up warm tones"
- "cozy cafe atmosphere"
Want me to adjust any of these?
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 258 lines · 57 tokens per session scan A ea722f9ea6aa
image-fetcher is a skill published in the GitHub repository sacredvoid/skillkit (10 stars, last pushed 5mo ago), licensed MIT. It adds 57 tokens to every session and 2,297 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
image-fetcher
Fetch relevant, high-quality, free-to-use images from the web. Accepts a description/query, or scans the current directory for context. Sources from Unsplash, Pexels, and Pixabay APIs (if keys configured) with a zero-config WebSearch fallback.
requirements-elicitation
Turns vague intent into testable requirements.
self-improvement
Captures lessons and promotes recurring patterns.
building-mcp-servers
Authors an MCP server with the official SDK and gates.
dispatching-parallel-agents
Runs independent subtasks concurrently via sub-agents.
security-audit
Detects high-confidence security risks in code.