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 houshuang/limbic --skill drivegit clone --depth 1 https://github.com/houshuang/limbicWrote 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/houshuang/limbic/drive)<a href="https://agentmods.dev/skills/houshuang/limbic/drive"><img src="https://agentmods.dev/badge/skills/houshuang/limbic/drive/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/houshuang/limbic/drive"><img src="https://agentmods.dev/badge/skills/houshuang/limbic/drive.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 | $0.00096 | $0.00692 |
| Opus 5.5 | $0.00038 | $0.00277 |
| Sonnet 5.5 | $0.00019 | $0.00138 |
| Haiku 4.5 | $0.00010 | $0.00069 |
Grade A, and why
drive 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 14d 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 — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Drive
Compile the request into the smallest pilot that can disprove a bad direction. Do not act as an autonomous project manager yet. The output is a direction card for the user to judge before execution begins.
Workflow
- Read the full request and the target project's instructions. Infer the desired user outcome, what "better" means, and the important non-goals. State useful assumptions instead of turning the opening into an interview.
- Before web research, broad code inspection, or implementation, retrieve the
nearest local analogs. Inspect at most three strong sources: recent project
artifacts, repository history, and prior conversations. If
claude-chat-searchis available, usecrossdirectly; do not re-index first. If no analog exists, record the queries andnone-found. - Choose
researchorimprove. Read only the matching mode guide: research or improve. - Propose one representative pilot of one to three units and the evidence that would tell the user whether it works. Prefer something the user can actually use, browse, inspect, or manually test.
- Stop before execution. In v0, set workers, model calls, and premium calls to zero; disable delegation; and forbid scaling until the user accepts the pilot.
- Draft a plan matching the plan contract. Check
it with
python -m limbic.drive validate PLAN.jsonwhen Limbic is importable. If it is not, apply the contract manually and say the automated check was unavailable. Do not install dependencies merely to run this check. - Present the concise direction card in chat. Ask at most one question, and only when two plausible answers would lead to materially different pilots.
Non-negotiable gates
- Retrieve before planning: local precedent precedes generic external advice.
- No batch larger than three until one complete representative unit has been experienced and accepted.
- No worker may spawn another worker. Future worker models must be explicit.
- Premium models are for convergence after uncertainty is visible, not for broad initial exploration.
- Stop on evidence: user acceptance, a failed usefulness test, a material change of direction, or a missing dependency that changes the plan.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 14d ago First seen · 55 lines · 96 tokens per session scan A 9a1701e49004
drive is a skill published in the GitHub repository houshuang/limbic (3 stars, last pushed 3d ago), licensed MIT. It adds 96 tokens to every session and 692 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 2026-09-17.
Other skills, from other repositories
sentence-transformers
Framework for state-of-the-art sentence, text, and image embeddings. Provides 5000+ pre-trained models for semantic similarity, clustering, and retrieval. Supports multilingual, domain-specific, and multimodal models. Use for generating embeddings for RAG, semantic search, or similarity tasks. Best for production…
cortex-distill
A skill for turning raw session records into organised notes and project documents. It uses a personal vault, which is a folder for storing and retrieving knowledge.
cortex-broadcast
Propagate insights from a distilled Raw into related existing Notes/Projects pages through conversational per-page edits.
cortex-query
A search tool for finding and retrieving information from Cortex, an external personal memory made up of stored notes and project pages.
cortex-evolve
A tool that saves technical notes and project decisions to a Cortex vault, which is an Obsidian note collection stored as files.
sentence-transformers
Framework for state-of-the-art sentence, text, and image embeddings. Provides 5000+ pre-trained models for semantic similarity, clustering, and retrieval. Supports multilingual, domain-specific, and multimodal models. Use for generating embeddings for RAG, semantic search, or similarity tasks. Best for production…