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/bumblebiber/its-over-9k/using-o9knpx skills add Bumblebiber/its-over-9k --skill using-o9kgit clone --depth 1 https://github.com/Bumblebiber/its-over-9kWrote 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/bumblebiber/its-over-9k/using-o9k)<a href="https://agentmods.dev/skills/bumblebiber/its-over-9k/using-o9k"><img src="https://agentmods.dev/badge/skills/bumblebiber/its-over-9k/using-o9k.svg" alt="Measured on agentmods" 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 | $0.00090 | $0.01361 |
| Opus 5 | $0.00045 | $0.00681 |
| Sonnet 5 | $0.00018 | $0.00272 |
| Haiku 4.5 | $0.00009 | $0.00136 |
Grade B, and why
using-o9k scanned grade B 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 5d 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.
Strips warnings and disclaimersmediumAnti-refusal
Omitting safety caveats hides risk from the user and is a common jailbreak preamble.
When an arbitration needs a human decision (e.g. two dispatch owners), don't lecture — surface it once and point to `/o9k-guide`, which walks the user How it starts
The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
using-o9k — The Doctrine
You are running under the o9k efficiency framework. Its goal: maximize the fraction of your context window doing useful work. Every token you emit or load gets re-read on every subsequent turn — waste compounds.
o9k itself owns no efficiency feature. It is the connector: it decides which installed framework owns which concern, wires them into every host, and keeps them from colliding. Whether a given rule below applies depends on what the user opted into — read the table, not your memory of a default install.
What applies, and when
| Situation | Concern | Owner (if installed) |
|---|---|---|
| Writing any response | Output compression | caveman (upstream project) |
| About to open/read/explore code | Context discipline | scout (o9k-scout) |
| A search, lookup, or decomposable task | Subagent isolation | dispatch path A (team-up) |
| External CLI worker (tmux / cross-CLI) | Multi-agent roster | team-up's dispatch path B — required once a roster exists |
| Session start, project questions, "what was the state?" | Memory | memory (o9k-memory) |
| Conflict between any of the above | Arbitration | this skill |
A row whose owner is not installed simply does not apply. Do not emulate a missing pillar by hand, and do not tell the user they are missing one unless they ask.
Roster detection gate (run before any external CLI spawn):
command -v team-up && test -f ~/.team-up/roster.json
- No roster → Path A only (in-host RESULT subagents).
- Roster present → external CLI workers must use Path B:
runs create→team-up dispatch --run-id→ cheap in-hostruns waitwatcher. Bare tmux dispatch without mailbox + watcher = incomplete spawn — parent never gets notified. Seedispatch§ Incomplete-spawn gate.
Core rules (always on)
- Never load what you can look up later. Prefer an ID, a path, a one-line summary over the full content. Drill down only when the task demands it.
- Never repeat what's already in context. No restating the user's request, no quoting code back unchanged, no summarizing your own previous message.
- Output is future input. Everything you say now is context tax on every
later turn. Say it once, say it short (see
caveman). - Isolate the noisy work. Broad searches, log dumps, doc reading — send a
subagent (see
dispatchpath A); keep only the conclusion. - External workers need a callback. When a team-up roster exists and
you spawn an external CLI in tmux, Path B is mandatory — not a shortcut,
not "when you remember." Three steps or the spawn is incomplete:
runs create,dispatch --run-id, watcher onruns wait. Never report "running" to the human until the watcher exists. - Persist before you lose it. Approaching compaction or
/clear: flush decisions, lessons, and open state to memory (seememory) — context is ephemeral, memory is not.
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.
- 5d ago First seen · 93 lines · 90 tokens per session scan B 3e1012f3a5e4
using-o9k is a skill published in the GitHub repository Bumblebiber/its-over-9k (0 stars, last pushed 8d ago), licensed MIT. It adds 90 tokens to every session and 1,361 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (strips warnings and disclaimers). 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.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…
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…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…