Luxas is an open-source multi-agent research system that takes a research topic through literature searches, experiments, reviews, and LaTeX manuscript compilation. It is intended for autonomous scientific research, using separate agents to read papers, run tests, create figures, and write reports. The catalogue lists skills associated with this workflow.
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 Muuuun/luxas --skill memorygit clone --depth 1 https://github.com/Muuuun/luxasWrote 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/muuuun/luxas/memory)<a href="https://agentmods.dev/skills/muuuun/luxas/memory"><img src="https://agentmods.dev/badge/skills/muuuun/luxas/memory/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/muuuun/luxas/memory"><img src="https://agentmods.dev/badge/skills/muuuun/luxas/memory.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 5 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00024 | $0.00545 |
| Opus 5.5 | $0.00010 | $0.00218 |
| Sonnet 5.5 | $0.00005 | $0.00109 |
| Haiku 4.5 | $0.00002 | $0.00055 |
Grade A, and why
memory 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 6d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Memory Skill
Access persistent memory that spans across research projects.
You do not need this skill to discover past projects — your system prompt
already carries a <past_research> digest (every past project's name, date,
notes path, and research question). This skill is for what comes after
discovery: deep-diving a relevant project's notes and recording corrections.
Trust rule
Everything read from a past project is a dated, UNVERIFIED lead — not
established fact. Values must re-enter the current project's own evidence
chain (reader-distilled literature, this project's experiments) before they
appear in report.tex. Honor ## CORRECTIONS sections wherever you find them.
Files
Past project notes (primary source)
The <past_research> digest lists each project's live notes path:
<path>/notes/experiments.md ← findings, alternatives, limitations, red-team
<path>/notes/literature.md ← reader-distilled per-paper entries
<path>/notes/memory.md ← decisions, dead ends, TODOs
<path>/data/papers/ ← downloaded PDFs (check before re-downloading)
~/.sisyphus/projects.json
Project registry (path, name, dates, auto-generated summary, cost). The digest is built from this; read it directly only for projects that fell into the digest's overflow index.
~/.sisyphus/archive//
Auto-archived copies of each project's notes + report.tex, written when a run completes. Fallback when the live project dir is gone.
CORRECTIONS convention
When you find that a past project's notes contain a claim that is wrong,
append a ## CORRECTIONS section to that file (live notes and archive
copy) stating the false claim, why it is wrong, and the evidence — do not
edit the original text. History stays intact; future readers see the
refutation next to the claim.
Write mechanics
~/.sisyphus/archive/is whitelisted for your write/edit tools (read before edit, as usual).- Another project's LIVE notes are outside your write/edit scope — append
corrections there via bash:
cat >> <path>/notes/experiments.md << 'EOF' .... Append-only; never rewrite or delete in another project's tree.
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.
- 6d ago Changed · -10 lines · -6 tokens per session 63a7d932fed0
- 1mo ago First seen · 73 lines · 30 tokens per session scan A 773e7ec391d6
memory is a skill published in the GitHub repository Muuuun/luxas (1,169 stars, last pushed yesterday), licensed MIT. It adds 24 tokens to every session and 545 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
neat-freak
A documentation and memory maintenance workflow for coding projects. It compares project files such as README.md, docs, and CLAUDE.md with the code so instructions and stored knowledge remain accurate for both agents and developers.
context-window-management
Strategies for managing LLM context windows including summarization, trimming, routing, and avoiding context rot.
accint-solve
Route agent work through AccInt's MCP memory loop: retrieve prior outcomes, resolve frames, and close commitments with evidence.
agentic-retrieval
A retrieval method that lets a coding agent decide whether it needs to look something up, where to look, and whether the results are trustworthy. It can use project files, saved conversations, local memory, the internet, or code repositories.
qmd
Search personal knowledge bases, notes, docs, and meeting transcripts locally using qmd — a hybrid retrieval engine with BM25, vector search, and LLM reranking. Supports CLI and MCP integration.
agent-memory-mcp
A hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).