A local tool for converting a folder of documents into a searchable knowledge graph, which maps people, topics, relationships, and facts. It can convert common document types to Markdown and optionally read scanned pages with OCR, or optical character recognition.

Use it to ingest files or folders, build document-based memory, and answer questions grounded in your own documents.

It lets an agent recall small relevant parts of your documents instead of repeatedly loading whole files and spending tokens reading them.

memorise — Skill for Claude Code by GRU-953: A local tool for converting a folder of documents…

Install

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.

agentmods

npx agentmods add skills/gru-953/mta_plugin/memorise

Any agent

npx skills add GRU-953/MTA_Plugin --skill memorise

Clone the repo

git clone --depth 1 https://github.com/GRU-953/MTA_Plugin

Made for: Claude Code.

Or install memorised-them-all, the plugin that ships this one along with the rest of its 1 skill, 5 commands, 1 MCP server.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/gru-953/mta_plugin/memorise.svg)](https://agentmods.dev/skills/gru-953/mta_plugin/memorise)

Your own site

<a href="https://agentmods.dev/skills/gru-953/mta_plugin/memorise"><img src="https://agentmods.dev/badge/skills/gru-953/mta_plugin/memorise.svg" alt="Measured on agentmods" height="20"></a>

Per session 106 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 739 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00106	$0.00739
Opus 5	$0.00053	$0.00369
Sonnet 5	$0.00021	$0.00148
Haiku 4.5	$0.00011	$0.00074

Measured 5d ago against content hash 2c8833aa702e, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

memorise 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 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.

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.

skills/memorise/SKILL.md · 48 lines

How it starts

The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Memorise them All

Turn any pile of documents into local, token-free graph memory for Claude, and recall from it cheaply. All conversion and digestion runs on the user's machine (MarkItDown, plus optional Tesseract OCR) — fully deterministic and model-free, no LLM/Ollama/GPU. Claude only ever issues a small tool call and gets back compact metadata or a tiny relevant slice — never whole documents.

When to use

The user wants to ingest/digest/memorise files, attachments, or a whole folder.
The user asks a question that should be answered from their own documents.
The user wants a knowledge graph of their documents.

How it works (the pipeline)

Convert — every attachment → Markdown locally (PDF/Office/HTML via MarkItDown, incl. legacy binary .doc/.ppt/.xls via optional LibreOffice; scanned images via optional Tesseract OCR; archives unpacked safely). Legacy Bengali typed in Bijoy/SutonnyMJ ANSI fonts is auto-upgraded to Unicode (font-aware, so mixed English+Bengali documents convert cleanly).
Segment → Extract — structure-aware chunks; rule-based, deterministic extraction of entities, relations, and atomic facts (no LLM/model).
Graph + themes — a knowledge graph with community-detected themes.
Layered memory — a global synopsis, per-theme summaries, per-document notes, and graph.json. Recall ranks with model-free BM25 (Bengali-aware) and returns a tiny cited slice, declining off-topic queries.

Tools

digest(paths, project?, reset?) — build/refresh memory (reset: true rebuilds from scratch). Returns metadata only.
convert(paths, out_dir?, project?) — convert files/dirs/globs to Markdown locally (legacy Bengali/SutonnyMJ → Unicode); writes .md files to out_dir (default markdown_converted/ beside the input). Token-free. Use when the user just wants Markdown, not a digest.
recall(query, project?, k?) — return a small, citable slice of memory.
memory_overview(project?) — synopsis + themes.
export_memory(dest, project?) — export portable Markdown files.
list_digestible(directory) — list convertible files (paths/sizes only).
forget(project?) — delete a project's memory (graph, converted Markdown, summaries/notes). Irreversible.
memory_status() — local stack health.

Read the full file on GitHub · 48 lines

Changes

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 · 48 lines · 106 tokens per session scan A 2c8833aa702e

Subscribe to this mod's changes

memorise is a skill published in the GitHub repository GRU-953/MTA_Plugin (1 stars, last pushed 1mo ago), licensed MIT. It adds 106 tokens to every session and 739 once invoked, about $0.0005 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.

Deterministic Memory and context bengali claude claude-desktop cursor gemini

agent-memory

Use the agent-memory MCP server to retrieve, verify, and store durable project knowledge and global user knowledge before and after non-trivial work.

RashadAnsari/myagents · 31 tokens

delx-mcp-server

Delx MCP Server. Prefer MCP tools if connected; otherwise the package CLI.

davidmosiah/delx-mcp-server · 22 tokens

skill

NodeSpace is context infrastructure for AI-native development — a local-first knowledge graph running on this machine.

NodeSpaceAI/nodespace-core · 0 tokens

recallium-guidance

Use when you have access to Recallium MCP tools. Provides guidance on memory storage, search, and knowledge management. CRITICAL - invoke the MCP tool recallium at session start, store memories after EVERY substantive interaction, and honor global rules.

recallium-ai/recallium · 57 tokens

karpathy-llm-wiki

Use when building or maintaining a personal LLM-powered knowledge base. Triggers: ingesting sources into a wiki, querying wiki knowledge, linting wiki quality, 'add to wiki', 'what do I know about', or any mention of 'LLM wiki' or 'Karpathy wiki'.

Astro-Han/karpathy-llm-wiki · 67 tokens

lc-curate-context

Decide which files a task actually needs, record that as a reusable llm-context rule, verify it against the codebase - including the files your selection references but leaves out - and pack it for your own context, a chat, or a sub-agent you dispatch. Load when choosing what code to put in front of a model, packing…

cyberchitta/llm-context.py · 90 tokens