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/a1f/agent-templates/pr-explainnpx skills add a1f/agent-templates --skill pr-explaingit clone --depth 1 https://github.com/a1f/agent-templatesWhat 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 | $0.00087 | $0.07251 |
| Opus 5 | $0.00044 | $0.03625 |
| Sonnet 5 | $0.00017 | $0.01450 |
| Haiku 4.5 | $0.00009 | $0.00725 |
Grade A, and why
pr-explain 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.
- **Safe commands only**: test runners, linters, builds, read-only CLI and `curl` against a How it starts
The opening of the file, as written. The whole thing — 437 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PR Explain
Write the page you would produce pair-programming with the person who owns this repo: you at the keyboard, walking them through the change you just made. They want to keep a clear mental model of code that agents typed. They are not hunting for bugs — a reviewer does that. They want to understand the whole change, not a sampler of it.
Three rules run through the whole page:
- Plain words. The
englishrule governs every word on the page — it loads each session, so you already have it. Say it the way you would to a smart friend who has not seen the code. If a code word is the only word that fits (hash,symlink,CI), say what it does in plain words right after — once. Talk to the reader: "you", "your". - The whole diff. Every change in the PR is either shown and explained or named in a roll-up line. A reader who finishes the walkthrough has seen the change — never two excerpts from a hundred-line diff.
- Proof is scored, not asserted. Output on the page is output you captured. Every piece of it sits on the ladder below, carrying its points. A reader sees at a glance whether this change was watched working or merely compiled. "All tests passed" is the weakest sentence on the page, never the proof.
Arguments
| Form | Meaning |
|---|---|
| (blank) | The open PR for the current branch |
#N or N |
PR number in the target repo |
| a PR URL | That PR |
--confirmed "<what you saw>" |
Fills the ladder's top row — you ran this change yourself. Quoted verbatim, dated today. Combines with any form above |
Runtime resolution
- Template:
~/.claude/at/templates/pr-explain-page.html— design tokens plus the.bottomline,.goals,.tree,.hunk-head,.testlist,.wit,.score,.rung,.short-proof,.gate-chip,.no-proof,.ba-label, and.cmdclasses the chapters use. Missing? Still ship: build a single-file HTML page with inline CSS that follows the chapter contract. Visual polish degrades; the contract does not. - Vale config:
~/.claude/at/templates/pr-explain.vale.ini, staged with the template. It names the write-good / proselint / Google packages the mechanical prose gate runs. Missing, orvalenot installed → the gate degrades to thewc+ banned-word checks. - Page file: write the filled page to the session scratch directory (never the target
repo) as
pr-<owner>-<repo>-<number>.html, owner/repo parsed from the PR'surl. - Target repo: the git repo of the working directory; run every
gh/gitcommand there.
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 · 437 lines · 0 tokens per session scan A 55f12fa5dfec
pr-explain is a skill published in the GitHub repository a1f/agent-templates (2 stars, last pushed 2d ago), licensed MIT. It adds 87 tokens to every session and 7,251 once invoked, about $0.0004 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…