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 Filipe-Soares-de-Almeida/MemAI --skill memai-distillgit clone --depth 1 https://github.com/Filipe-Soares-de-Almeida/MemAIWrote 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/filipe-soares-de-almeida/memai/memai-distill)<a href="https://agentmods.dev/skills/filipe-soares-de-almeida/memai/memai-distill"><img src="https://agentmods.dev/badge/skills/filipe-soares-de-almeida/memai/memai-distill/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/filipe-soares-de-almeida/memai/memai-distill"><img src="https://agentmods.dev/badge/skills/filipe-soares-de-almeida/memai/memai-distill.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 | $0.00141 | $0.02419 |
| Opus 5 | $0.00071 | $0.01210 |
| Sonnet 5 | $0.00028 | $0.00484 |
| Haiku 4.5 | $0.00014 | $0.00242 |
Grade A, and why
memai-distill 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 11d 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 — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
memai-distill — distill / move type / merge
Saves the durable knowledge that is stuck in the wrong type, or held in a checkpoint that is about to be archived, without losing any of it.
The type boundary ([[memai-memory]] §1):
note= a timeless fact;reasoning= an analysis trace;anti_pattern= a trap;checkpoint= transient bearing. When a checkpoint (or areasoning) carries a fact that outlives the session, that fact wants to be anote.
1. distill — extract the durable (ARCHIVES the sources)
optimize_stage kind distill, payload {source_uids:[...], new_type:"note|reasoning|anti_pattern", new_content, title, tags?, domain?}:
creates the new memory, links supersedes from it to every source and
archives every source. The dashboard's undo purges the created memory
and restores each source, so the whole thing is reversible. Two uses:
- retiring the checkpoints of finished work while keeping what they taught;
- n-ary merge: folding N near-duplicates into one, with synthesized content.
verified is required — the kind archives its sources. Say what the
live-facts check was and where the durable content landed.
Payload details:
new_typeaccepts onlynote/reasoning/anti_pattern.checkpoint/handoffare not targets, and neither isdiagram: a flow is written bydiagram()+diagram_node/diagram_edge, never from prose.reasoningandanti_patternare made of FIELDS, so anew_contentfor either has to read back as those fields — each label opening its own line, in order, none empty:HYPOTHESIS: … / REASONING: … / RESULT: … / REVISED BELIEF: … / NEXT TIME: …andTEMPTATION: … / WHY WRONG: … / INSTEAD: …. A body that does not is refused at staging and comes back inerrors, naming the field. Distilling intonotehas no such shape.domainis a path (acme/x100/p200). The new memory inherits nothing from its sources, so state where it is filed — omit it and the memory is filed nowhere. A cross-cutting axis does not go here: after the apply,crosslist({"also": ["omni/x900"]}) oralso_domain(uid, 'omni/x900').- The payload accepts
source_uids,new_type,new_content,title,tagsanddomain, and nothing else. Any other key is refused at staging and returned inerrors, soalso,review_after,source_refandsessiondo not travel with a distill — they are the follow-up in §2. titleis required, at most 120 characters.distillis the only suggestion kind that AUTHORS a memory, and nothing names it afterwards: staged without one, the new memory would be listed by the opening line of its body forever. Name it by its subject, in the words someone would look for it by — the same line anote()would have taken.- A
diagramcannot be a source. Its content is the projection of its graph, and archiving it trades a navigable flow for prose, so staging refuses it. An obsolete diagram is an explicitarchive; a wrong one is a fix to the graph. It never arrives as a candidate fromdedup_scaneither, which excludes diagrams from its pairs.
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.
- 11d ago First seen · 180 lines · 141 tokens per session scan A b6073293e618
memai-distill is a skill published in the GitHub repository Filipe-Soares-de-Almeida/MemAI (0 stars, last pushed yesterday), licensed MIT. It adds 141 tokens to every session and 2,419 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
browserwing-executor
Control browser automation through HTTP API. Supports page navigation, element interaction (click, type, select), data extraction, accessibility snapshot analysis, screenshot, JavaScript execution, and batch operations.
dev-browser
Browser automation with persistent page state. Use when users ask to navigate websites, fill forms, take screenshots, extract web data, test web apps, or automate browser workflows. Trigger phrases include "go to [url]", "click on", "fill out the form", "take a screenshot", "scrape", "automate", "test the website"…
ask-user-question
Ask users questions via the UI. Use when you need clarification, user preferences, or confirmation before proceeding. The user CANNOT see CLI output - this tool is the ONLY way to communicate with them.
memos-memory-guide
Use the MemOS Local memory system to search and use the user's past conversations. Use this skill whenever the user refers to past chats, their own preferences or history, or when you need to answer from prior context. When auto-recall returns nothing (long or unclear user query), generate your own short search query…
memmachine-memory
Use when an agent or model needs durable project, user, or session context from MemMachine, needs to save information to MemMachine memory, has requests involving mem-cli, memmachine, or memmachineclient, has insufficient conversation context, or is tempted to search local files for prior context that should come from…
browserwing-admin
Manage and operate BrowserWing — an intelligent browser automation platform. Install dependencies, configure LLM, create/manage/execute automation scripts, use AI-driven exploration to generate scripts, browse the script marketplace, and troubleshoot issues.