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/vaayne/agent-kit/curatornpx skills add vaayne/agent-kit --skill curatorgit clone --depth 1 https://github.com/vaayne/agent-kitWhat 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 | $0.00077 | $0.00634 |
| Opus 5 | $0.00039 | $0.00317 |
| Sonnet 5 | $0.00015 | $0.00127 |
| Haiku 4.5 | $0.00008 | $0.00063 |
Grade A, and why
curator 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 — 39 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Curator
Harvest writes; curator keeps it trustworthy. An unmaintained knowledge base rots — contradictions accumulate, duplicates fragment retrieval, and references outlive the things they point to. Two passes, run together or alone.
Lint pass
Routine-tier work: mechanical checks, conservative fixes.
- Scope the window. Find the last curator-run memory (
nmem memories search "curator run"); lint what changed since. First run: sweep everything (nmem memories list,nmem stats). - Duplicates — search topic clusters the window touched; near-identical entries merge into the strongest one (
update), losers getsupersedelinks. - Contradictions — where a newer memory disagrees with an older one, verify which is true now;
deprecatethe loser with the reason recorded, never silently. - Dead references — memories naming files, flags, skills, or commands: verify each still exists (check the repo or filesystem). Update what moved; deprecate what vanished.
- Expired — entries whose stated expiry has passed: re-verify or deprecate. Time-sensitive claims with no expiry get one added now.
- Harvest gaps — list the window's threads (
nmem threads list) and compare against memories added in the same span; a non-trivial session that saved nothing is a harvest miss. Read the thread, save what a future session would need, or confirm there was genuinely nothing to keep. Threads import is manual — a sparse window may mean unsynced, not idle; say so in the report instead of guessing.
Fix verbs, in order of preference: update (correct in place) → supersede (old links to replacement) → deprecate (obsolete, kept for audit) → delete (only for entries wrong from birth). History stays legible.
Synthesis pass
Premium-tier work: judgment across the whole window.
Read the window's memories plus the feed (nmem feed), then write a short report to the user: what entered, what themes are forming, what drifts from earlier decisions, what deserves attention next. This is the pass that surfaces cross-source connections no single entry shows.
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 · 39 lines · 77 tokens per session scan A 8aaa34ed4a14
curator is a skill published in the GitHub repository vaayne/agent-kit (53 stars, last pushed 5d ago), licensed MIT. It adds 77 tokens to every session and 634 once invoked, about $0.0004 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…