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 commands/erhangundogan/videre/watchgit clone --depth 1 https://github.com/erhangundogan/videreWrote 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/commands/erhangundogan/videre/watch)<a href="https://agentmods.dev/commands/erhangundogan/videre/watch"><img src="https://agentmods.dev/badge/commands/erhangundogan/videre/watch.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00013 | $0.01901 |
| Opus 5 | $0.00006 | $0.00950 |
| Sonnet 5 | $0.00003 | $0.00380 |
| Haiku 4.5 | $0.00001 | $0.00190 |
Grade A, and why
watch 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 4d 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 — 187 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A long-running loop that keeps the pipeline populated, so readers always see fresh data without you rerunning things by hand. No server and no UI: it runs in the foreground logging to stderr until you stop it with Ctrl-C.
videre watch ~/Photos # scan, faces, HEIC cache, and locations every 5 minutes
videre watch # same, using the folder from `videre config set path`
videre watch ~/Photos --scan --faces # only these stages
videre watch ~/Photos --heic # only pre-convert HEIC thumbnails
videre watch ~/Photos --location # only look up place names
videre watch ~/Photos --prune # also clean stale entries (off by default)
videre watch ~/Photos --interval 60 # seconds between cycles (default 300)
videre watch ~/Photos --silent # no per-cycle output
videre watch ~/Photos --db ~/photos.db # use a specific database
videre watch ~/Photos --type image # only watch for new images
videre watch ~/Photos --path ~/Photos/Inbox # only watch one subfolder
:::tip These filters work the same way across commands, and combine. See scoping a run. :::
--output-sqlite still works as an alias for --db, the name it had
originally. Existing scripts do not need changing.
Stages
If none of --scan, --faces, --heic or --location are given, all four
run. --prune and --export-xmp are the exceptions: they are opt-in and never
default on.
| Stage | What it does |
|---|---|
--scan |
Same scan and hash pipeline as videre scan, including reading marks from XMP (--xmp db|file|newest, see scan) |
--faces |
Detects faces in new images, then regroups everything |
--heic |
Pre-converts and caches HEIC thumbnails |
--location |
Looks up place names for GPS coordinates that have none |
--prune |
Same cleanup as videre prune |
--export-xmp |
Writes labels to .xmp sidecars, same as videre export |
The export stage keeps sidecars current while you work in another tool. It is
opt-in per run with --export-xmp, or always-on by setting
videre config set export-xmp-on-watch true. It merges into existing sidecars,
so it never clobbers another tool's data.
Note that embed and classify are
not stages. Semantic search data is not kept current automatically; run
videre embed yourself after adding a batch of photos.
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.
- 4d ago First seen · 187 lines · 13 tokens per session scan A d20b21668d7f
watch is a command published in the GitHub repository erhangundogan/videre (2 stars, last pushed 6d ago), licensed Apache-2.0. It adds 13 tokens to every session and 1,901 once invoked, about $0.0001 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 2026-08-31.
Other commands, from other repositories
deslop
Check the diff against main, and remove all AI generated slop introduced in this branch.
contrib
Profile how top GitHub contributors work — ingest, build personas, and synthesise a capability model + learnable playbook.
memory-init
Initialize local memory store + vector index for this project.
graph
Generate an interactive visual graph of your memories. Powered by graphify (github.com/safishamsi/graphify, MIT).
git
Git operations with intelligent commit messages and workflow optimization.
checklist
Generate a custom checklist for the current feature based on user requirements.