Borrowing it
Nothing to install: this file belongs to lucasmella-stack/misMEM. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/lucasmella-stack/misMEM/main/AGENTS.mdgit clone --depth 1 https://github.com/lucasmella-stack/misMEMWrote 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/instructions/lucasmella-stack/mismem/agents-md)<a href="https://agentmods.dev/instructions/lucasmella-stack/mismem/agents-md"><img src="https://agentmods.dev/badge/instructions/lucasmella-stack/mismem/agents-md/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/instructions/lucasmella-stack/mismem/agents-md"><img src="https://agentmods.dev/badge/instructions/lucasmella-stack/mismem/agents-md.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 | $0.04437 | $0.04437 |
| Opus 5.5 | $0.01775 | $0.01775 |
| Sonnet 5.5 | $0.00887 | $0.00887 |
| Haiku 4.5 | $0.00444 | $0.00444 |
Grade A, and why
misMEM AGENTS.md 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 yesterday.
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 — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Contrato de trabajo para cualquier agente (Claude Code a través de CLAUDE.md, que sólo importa este
fichero; Copilot; opencode) y para quien contribuya. Si este fichero y el código discrepan, gana el código:
se corrige el fichero.
Qué es misMEM
Memoria persistente para LLMs, expuesta como servidor MCP sobre una sola base SQLite que comparten todos los
clientes. Tiene tres capas: episodes (texto literal y fechado), memories (resúmenes que destila una
consolidación con LLM, con una saliencia que decae y se refuerza con el uso) y traits (patrones
cristalizados). Seis herramientas canónicas (capture, recall, reinforce, consolidate,
crystallize, forget) y alias mem_* de Engram. El resto de la forma del sistema:
docs/architecture.md.
Stack: TypeScript estricto, ESM, Node 22 o superior, better-sqlite3 (WAL y FTS5), SDK de MCP, Zod y Vitest;
la interfaz es React, Vite y Tailwind. Gestor de paquetes: pnpm, nunca npm.
Dónde vive qué
| Ruta | Qué posee | Se toca cuando |
|---|---|---|
engine/src/db.ts |
SCHEMA, migraciones aditivas (migrate()), apertura de la base con permisos 0600 |
cambia una tabla, una columna o un índice |
engine/src/schema.ts |
Inputs Zod de las herramientas y el tipo RecallHit |
cambia lo que una herramienta acepta o devuelve |
engine/src/tools.ts |
Las operaciones MCP: capture, recall y recallHybrid, reinforce, consolidate, crystallize, forget y la lista de episodios pendientes |
cambia lo que hace una herramienta |
engine/src/search.ts |
Ranking del recall: consulta FTS5, rescate y fusión RRF con la saliencia | cambia el ranking (pide mismem-bench) |
engine/src/embeddings/ |
provider.ts (configuración por entorno y cliente HTTP), store.ts (SQL de memory_embeddings, búsqueda y backfill), vector.ts (coseno) |
cambia la capa semántica (pide mismem-bench) |
engine/src/salience.ts |
Decay y refuerzo de la saliencia | cambia la fórmula de saliencia |
engine/src/noise.ts, engine/src/terms.ts, engine/src/hash.ts |
Marcadores de ruido, tokenización compartida por la búsqueda y la medición de uso, hash del contenido (dedup de capture) |
cambia uno de esos tres criterios |
engine/src/relations.ts |
Lecturas que comparten el visor y la exportación a Obsidian (memorias parecidas por vector) | cambia cómo se relacionan las memorias |
engine/src/server-core.ts |
Descriptores de las herramientas MCP y su despacho; lo usan server.ts (stdio) y server-http.ts |
se añade o cambia una herramienta |
engine/src/engram-compat.ts |
Los alias mem_* y su mapeo a las seis herramientas |
cambia la compatibilidad con Engram |
engine/src/server-http.ts |
Servidor HTTP: /mcp, /healthz, /viewer y /api/* |
se añade una ruta; la lógica va a otro módulo |
engine/src/http-guard.ts, engine/src/http-host.ts, engine/src/http-body.ts |
Dirección de enlace y BasicAuth, guarda de Host, lectura del cuerpo con tope |
cambia qué petición se atiende |
engine/src/viewer.ts, engine/src/viewer-api.ts, engine/src/viewer-actions.ts, engine/src/viewer-static.ts |
API de lectura del visor, sus escrituras por POST, y la interfaz compilada con su CSP | cambia lo que el visor muestra o puede hacer |
engine/ui/ |
La interfaz (React) en ui/src; se compila a dist/ui. ui/src/components/ui/ es código de shadcn: Biome y knip lo tratan aparte |
cambia lo que se ve |
engine/src/consolidation/ |
run.ts (la corrida), llm.ts (proveedores), prompts.ts, job.ts (corrida en segundo plano desde la interfaz) y su README.md |
cambia cómo se destilan los episodios |
engine/src/pareto/classifier.ts |
Prefiltro heurístico que descarta ruido antes del LLM | cambia qué se descarta sin LLM |
engine/src/ingest/ |
Importadores de ChatGPT y de VS Code, y el recorrido de ficheros | se importa un formato nuevo |
engine/src/inject.ts, engine/src/usage.ts, engine/src/hook-client.ts, engine/src/hook-readiness.ts |
Lógica de los hooks: qué inyectar, qué se usó, cliente remoto y espera a que el servidor esté listo | cambia un hook |
engine/src/obsidian.ts, engine/src/bench.ts |
Exportación como bóveda de Obsidian; medición del recall | cambia una de las dos |
engine/src/*-cli.ts, engine/src/*-args.ts, engine/src/cli-args.ts, engine/src/env.ts |
Los binarios mismem-*, la lectura estricta de sus argumentos y la lectura de variables numéricas |
cambia un CLI o un flag |
engine/tests/ |
Un <área>.test.ts por área; fixtures/ (exports ficticios) y support/ |
cada cambio |
scripts/ |
install.sh (instalador de Linux), mismem-hook.sh (envoltorio de los hooks), backup-mismem-from-server.ps1 (respaldo desde Windows) |
cambia la instalación o el respaldo |
engine/Dockerfile.mismem-brain, .github/ |
La imagen, los workflows, las plantillas y las etiquetas | cambia el empaquetado o la CI |
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.
- yesterday First seen · 168 lines · 4,437 tokens per session scan A 9e53a9e7f58c
misMEM AGENTS.md is an instructions file published in the GitHub repository lucasmella-stack/misMEM (3 stars, last pushed today), licensed MIT. It adds 4,437 tokens to every session, about $0.0177 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 2026-10-29.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.