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 umputun/agterm --skill annotate-documentgit clone --depth 1 https://github.com/umputun/agtermWrote 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/umputun/agterm/annotate-document)<a href="https://agentmods.dev/skills/umputun/agterm/annotate-document"><img src="https://agentmods.dev/badge/skills/umputun/agterm/annotate-document/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/umputun/agterm/annotate-document"><img src="https://agentmods.dev/badge/skills/umputun/agterm/annotate-document.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00142 | $0.01975 |
| Opus 5.5 | $0.00057 | $0.00790 |
| Sonnet 5.5 | $0.00028 | $0.00395 |
| Haiku 4.5 | $0.00014 | $0.00198 |
Grade A, and why
webrev scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
webrev
Requires agterm (AGTERM_SESSION_ID set).
Launch
Create a backup directory first, then run it as a foreground Bash call with timeout: 600000. A
background task lets Claude Code's status hooks overwrite the blinking blocked status the page sets. A
review that outlasts the cap is recovered from the backup file:
mktemp -d /tmp/webrev-XXXXXX
"${CLAUDE_SKILL_DIR}/scripts/webrev.py" review <file> --output <that dir>/result
The backup path must not exist beforehand: a missing file means no outcome was recorded, while an empty one means a submit with no notes.
The page opens as a floating panel over this session and the script waits for it, at --size N percent
(default 95) for documents, plans and anything long. Pass --fit instead for a short draft such as a commit
message, a comment or a short PR body: the panel opens at a size estimated from the text.
A plain-text draft gets 80 columns; markdown gets its longest line within 60..100 columns, and the height
follows the wrapped rows. agterm limits each side to 80% of the session, and anything taller, a wide table or
an opened note scrolls. The estimate is for agterm's HTML page zoom of 1.0; agterm reports the zoom only once
a page is open, so at any other zoom the panel resizes once right after it opens, to the estimate times the
zoom. A zoom changed while the page is up does not resize it. --pane left|right puts the page over one
split pane, always at full size, so neither --size nor --fit applies.
If --fit fails because the session is presented from another Mac (--width/--height are not supported),
run the review again without it.
WEBREV_STYLE=terminal in the environment switches the page to a flat terminal-like look: body text at one
size, square chrome, a status bar at the bottom. Unset keeps the default look; any other value is an error.
When the file is a temp file holding generated content, pass --title "<what it is>", for example
--title "PR description: add retries": the header shows it in place of the temp file name. Annotation
records keep the real file path.
What ships with it
8 files 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.
- 3d ago First seen · 124 lines · 142 tokens per session scan A 7ff8a462a17d
webrev is a skill published in the GitHub repository umputun/agterm (709 stars, last pushed today), licensed MIT. It adds 142 tokens to every session and 1,975 once invoked, about $0.0006 per session on Opus 5.5. 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 2026-10-25.
Other skills, from other repositories
cmd-review
Review the current diff or selected files (usage: /review [instructions | --last-diff | --target | --file | files...] [--style ]).
patch-edit
Apply one rewrite-plan step at a time as small reviewable file edits, never rewriting whole files when a localised change suffices.
m15-anti-pattern
Use when reviewing code for anti-patterns. Keywords: anti-pattern, common mistake, pitfall, code smell, bad practice, code review, is this an anti-pattern, better way to do this, common mistake to avoid, why is this bad, idiomatic way, beginner mistake, fighting borrow checker, clone everywhere, unwrap in production…
coding-guidelines
Guidance for Rust code style and best practices, including naming, formatting, comments, linting, and code review.
diff-review
Render the patch-edit run's accumulated changes as a reviewable diff, surface it through a GenUI choice surface, and persist the user's accept / reject decision into the artifact manifest.
orbit-github
Open Orbit briefing skill — selected by the Orbit pipeline when GitHub is the user's only connected connector, or when the user explicitly scopes their daily digest to GitHub. Pulls the past 24 hours of PRs, review requests, issues, CI runs, and merges from the user's authenticated GitHub connection and renders them…