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 skills add drolosoft/immich-photo-manager --skill photo-searchgit clone --depth 1 https://github.com/drolosoft/immich-photo-managerWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/drolosoft/immich-photo-manager/photo-search)<a href="https://agentmods.dev/skills/drolosoft/immich-photo-manager/photo-search"><img src="https://agentmods.dev/badge/skills/drolosoft/immich-photo-manager/photo-search/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/drolosoft/immich-photo-manager/photo-search"><img src="https://agentmods.dev/badge/skills/drolosoft/immich-photo-manager/photo-search.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What 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 5.1 | $0.00109 | $0.03823 |
| Opus 5 | $0.00055 | $0.01912 |
| Sonnet 5 | $0.00022 | $0.00765 |
| Haiku 4.5 | $0.00011 | $0.00382 |
Grade A, and why
photo-search scanned grade A with 0 findings 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 8d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 288 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Photo Search
Connection Required: ALWAYS CHECK FIRST
Before doing ANYTHING else in this skill, call ping on the Immich MCP server.
- If
pingsucceeds -> proceed with the skill normally. - If
pingfails or the MCP tools are not available -> STOP. Do not continue. Tell the user:
Immich is not connected. This plugin needs a running Immich MCP server to work.
Run /setup-immich-photo-manager to configure your Immich connection. You'll need:
- Your Immich server URL (e.g.,
http://192.168.1.100:2283)- An Immich API key (how to create one)
- The MCP server configured (see /setup-immich-photo-manager)
Do NOT skip this check. Do NOT try to run any other tool first. Always ping, always block if it fails.
CRITICAL RULE: NEVER CREATE TEMPORARY ALBUMS
This is the most important rule of this skill. NEVER create albums as part of a search workflow.
The user has a curated library of real albums. Creating temporary albums pollutes their library with junk. Instead:
- If photos belong to a real album -> use that real album directly
- If photos are NOT in any album -> show them directly using
get_thumbnails_batch - NEVER call
create_albumfrom this skill. Not for "temp" albums, not for "search results", not for any reason.
Search Workflow (Step by Step)
Step 1: Parse user intent
Identify what the user is looking for. Determine which search dimensions apply.
If the user is browsing rather than looking for something specific ("what's in here?", "surprise me"), start with search_explore for one representative photo per city and per detected concept, or search_random for a quick sample. Both answer in one call.
Step 2: Search for matching photos
Use search_metadata and/or search_smart to find matching assets.
IMPORTANT: how Immich stores place names:
- Immich stores cities as municipalities, not tourist names. "Tikal" does not exist as a city; it's in the municipality of "Flores" (state: "Petén", country: "Guatemala").
- "Lanzarote" does not exist as a city; look for municipalities like "Arrecife", "Yaiza", "Teguise", "Haría", etc. (state: "Canary Islands", country: "Spain").
- When a place-name search returns 0 results, try broader terms: search by state or country instead of city, then filter. Or use
search_smartwith the place name as a CLIP query. - Do not guess spellings.
search_citieslists every city that appears in the library, andsearch_suggestions(suggestion_type="city", country="Spain")returns the exact stringssearch_metadataexpects. One call is cheaper than three failed searches, and the same works forcamera-makeandcamera-model.
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.
- 8d ago Changed 2da3dbca32f3
- 8d ago Changed · +9 lines f2015994e6d7
- 12d ago First seen · 279 lines · 109 tokens per session scan A 54c26f78e650
photo-search is a skill published in the GitHub repository drolosoft/immich-photo-manager (47 stars, last pushed 8d ago), licensed MIT. It adds 109 tokens to every session and 3,823 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
ltxv2-video
Build Lightricks LTX-2 / LTX-2.3 video workflows covering text-to-video, image-to-video, GGUF and bundled checkpoints, distilled model, camera control LoRAs, synchronized audio, two-stage upscaling, and swapping alternate/GGUF base models.
ideogram-ultra
Build Ideogram 4 (Ideogram Ultra) txt2img and img2img workflows with the local open-weights model, dual conditional/unconditional models with DualModelGuider, Qwen3-VL text encoder, and structured JSON ("compositional deconstruction") prompts for strong text rendering and layout control.
video-extend
Extend / continue a video temporally with Pusa 2.2 in ComfyUI. Temporal flowmatching (the flowmatchpusa scheduler + WanVideoAddPusaNoise) on the WanVideoWrapper stack with WAN 2.2 T2V A14B (HIGH/LOW) models and the Pusa V1 LoRAs, conditioning on the loaded clip via WanVideoEncode so the existing motion carries into…
director
Full production pipeline covering story to scenes, Z-Image start frames, Qwen Edit end frames, WAN FLF video clips, ffmpeg concatenation.
ernie-image
Build Baidu ERNIE-Image / ERNIE-Image-Turbo workflows, primarily TEXT-TO-IMAGE. Pick ERNIE when you need precise multilingual text rendering, posters/signage, manga/anime multi-panel layouts, or strong instruction following for complex multi-object scenes. Also supports denoise-based image-to-image refine (NOT…
model-compatibility
Model family compatibility matrix covering loaders, resolutions, samplers, CFG, VAE, ControlNet, and LoRA compatibility for SD 1.5, SDXL, Flux, SD3, and video models.