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/pubky/pubky-app/code-reviewnpx skills add pubky/pubky-app --skill code-reviewgit clone --depth 1 https://github.com/pubky/pubky-appWhat 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 | $0.00033 | $0.02405 |
| Opus 5 | $0.00016 | $0.01203 |
| Sonnet 5 | $0.00007 | $0.00481 |
| Haiku 4.5 | $0.00003 | $0.00241 |
Grade A, and why
code-review 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 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.
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 — 254 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Local Code Review
Review code changes against project standards using parallel focused reviewers.
Presentation
Do NOT narrate your steps, show tool outputs, or explain what you are reading. After the user picks an option, immediately output this single message and nothing else until the review is ready:
Robots are reviewing your code...
Then silently perform steps 2-5 using tool calls. Do not output any text between the waiting message and the final report. When the review is complete, replace the status with the full report from step 6.
Step 1 — Ask the user
Before doing anything, ask:
What do you want to review?
- Local uncommitted changes
- Full branch diff against
dev- Specific files (provide paths)
Wait for the answer. If the user picks 3, ask them for the file paths.
Step 2 — Get the diff
- If 1: run
git diffandgit diff --staged - If 2: run
git diff origin/dev...HEAD - If 3: run
git diff -- <paths>and read the specified files
Store the full diff text — you will pass it to sub-agents.
Step 3 — Load project context
Read these files (you will pass their contents to the Architecture agent):
.greptile/config.json— structured rule list with scopes and severities.greptile/rules.md— detailed explanations and examples
Step 4 — Dispatch parallel review agents
Launch three Task sub-agents in parallel (all with readonly: true). Each agent receives the diff and a focused brief. Send all three in a single message so they run concurrently.
Agent 1: Architecture (project-specific)
Use subagent_type: "code-reviewer".
Prompt template (fill in {diff}, {config_json}, {rules_md}):
You are reviewing code changes against this project's architecture rules.
## Rules
{config_json}
## Rule Details
{rules_md}
## Diff to review
{diff}
## Instructions
Check every changed file against the rules above. For each rule, check if the diff introduces a violation within that rule's scope.
VERIFICATION IS MANDATORY. For every potential issue:
- Read the actual source file to confirm the violation exists
- If a rule references a layer boundary, verify import paths and call sites in the file
- If flagging an unused export, search the codebase to confirm it is truly unused
- If flagging dead code, verify the condition is actually always true/false
Never report anything you have not confirmed in source. Never use "likely", "probably", or "appears to".
For deeper context, consult the relevant docs/ file (see docs/README.md for the index) and docs/adr/ for Architecture Decision Records.
## Output format
Return ONLY a JSON array. No markdown, no explanation outside the array.
Each finding:
{
"file": "path/to/file.ts",
"line": 42,
"rule_id": "no-service-from-controller",
"severity": "high",
"what": "PostController imports LocalPostService directly, bypassing Application layer",
"why": "Violates layer boundary (docs/architecture.md, ADR-0004). Controllers must go through Application for all IO.",
"fix": "Replace LocalPostService.create(post) with PostApplication.commitCreate({ compositePostId, post, ... })"
}
If no violations found, return an empty array: []
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 · 254 lines · 33 tokens per session scan A 9dd4fd41f603
code-review is a skill published in the GitHub repository pubky/pubky-app (23 stars, last pushed 4d ago), licensed MIT. It adds 33 tokens to every session and 2,405 once invoked, about $0.0002 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
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.
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…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…