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 rjroy/vibe-garden --skill stratifygit clone --depth 1 https://github.com/rjroy/vibe-gardenWrote 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/rjroy/vibe-garden/stratify)<a href="https://agentmods.dev/skills/rjroy/vibe-garden/stratify"><img src="https://agentmods.dev/badge/skills/rjroy/vibe-garden/stratify/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/rjroy/vibe-garden/stratify"><img src="https://agentmods.dev/badge/skills/rjroy/vibe-garden/stratify.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 | $0.00088 | $0.00992 |
| Opus 5.5 | $0.00035 | $0.00397 |
| Sonnet 5.5 | $0.00018 | $0.00198 |
| Haiku 4.5 | $0.00009 | $0.00099 |
Grade A, and why
stratify 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 1mo 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 — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Stratify
Reorganize .lore/reference/ into topical category directories once the flat layout stops scanning well. Move pages, keep every link working, and keep the index authoritative.
Stratify changes where pages live, never what they say. Page bodies, frontmatter, fg-type, fg-sources, and fg-evidence all stay untouched except for link path repairs.
Thresholds
- A directory earns stratification when it holds more than ~12 pages (
index.*and dotfiles don't count). - Split an overgrown directory into 3-4 groups. As the wiki grows, adjust toward 6-7 top-level categories.
- Never create a directory that would hold fewer than ~3 pages; leave those pages where they are until siblings accumulate.
- Apply the rule recursively: a category gets subcategories only when it exceeds the threshold itself. Categories under the threshold stay flat.
- After the first stratification, later runs are incremental: split only the directories that outgrew the threshold. Don't reshuffle pages that already fit or rename categories without cause — every move has a link-repair cost.
Steps
1. Inventory and design the taxonomy.
Read the index (.lore/reference/index.md, or index.html if that's all there is) and list every page on disk, recursively. Pages may be .md or .html; both move the same way. Note any page on disk the index doesn't list — those are orphans; adopt them into the new index during this run rather than leaving them invisible.
Group pages by topic, not by fg-type — type is already recorded in each page's frontmatter, and a category mixing decisions, lessons, and architecture about the same subject is the point. Derive topics from titles, index descriptions, and tags; skim page bodies only where those are ambiguous. Use short kebab-case directory names.
Present the proposed mapping as a table (category, subcategories, page count) before moving anything, and adjust if the user objects.
2. Survey referrers.
Before moving, find everything that links to pages about to move:
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.
- 1mo ago First seen · 66 lines · 88 tokens per session scan A 57c1a5045a54
stratify is a skill published in the GitHub repository rjroy/vibe-garden (6 stars, last pushed 18d ago), licensed MIT. It adds 88 tokens to every session and 992 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
pace-learn
A knowledge-management process that records reusable lessons and patterns from completed reviews, failed checks, rejected work, and signs of struggle.
pace-pulse
Auto-invoked during advance mode after extended work on same CR, at session start/end, or when rhythm anomalies are detected. NOT for metrics reports or retrospectives (use /pace-retro). NOT for project status overview (use /pace-status).
add-teams
Adds Microsoft Teams connector to a Power Apps code app. Use when sending Teams messages, posting to channels, or integrating with Teams chat.
sys-configure
Configure Claude Octopus — redirects to /octo:setup interactive wizard.
relations
OPS on-demand: This skill should be used when the user asks to "who do I owe a reply", "relationship…
llm-friendly-context
Clarifies inputs, outputs, success criteria, decisions, and unresolved conditions so downstream consumers can execute without guessing. Use when writing or revising LLM-facing prompts, handoffs, planning artifacts, reviews, reports, or generated instructions.