qmemory

qmemory is a skill for Claude Code, Codex from 3awny/qship. It costs 64 tokens per session (1,827 once invoked), scanned A, original, MIT.

A guide for saving reusable lessons from a task into persistent project memory, so they can be found in later sessions.

In plain words
What is it for?
Use it to review completed work and record durable conventions, fixes, and non-obvious lessons.
Why use it?
It prevents important corrections, discoveries, and recurring pitfalls from being lost when the conversation ends.

Skill for Claude CodeCodex

Part of the qship plugin — 22 skills, 1 agent, 1 hook shipped together

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/3awny/qship/qmemory
Any agent
npx skills add 3awny/qship --skill qmemory
Clone the repo
git clone --depth 1 https://github.com/3awny/qship

Made for: Claude Code, Codex.

Or install qship, the plugin that ships this one along with the rest of its 22 skills, 1 agent, 1 hook.

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.

agentmods badge for qmemory

README.md
[![agentmods](https://agentmods.dev/badge/skills/3awny/qship/qmemory.svg)](https://agentmods.dev/skills/3awny/qship/qmemory)
Your own site
<a href="https://agentmods.dev/skills/3awny/qship/qmemory"><img src="https://agentmods.dev/badge/skills/3awny/qship/qmemory.svg" alt="Measured on agentmods" height="20"></a>
Per session 64 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,827 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.

ModelPer sessionOnce invoked
Fable 5 $0.00064 $0.01827
Opus 5 $0.00032 $0.00914
Sonnet 5 $0.00013 $0.00365
Haiku 4.5 $0.00006 $0.00183

Measured 3d ago against content hash 21d3b076239e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

qmemory 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 3d 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.

templates/skills/qmemory/SKILL.md · 145 lines

How it starts

The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Knowledge Item Memory Storage

Review the current conversation and extract generic, reusable lessons worth carrying into future sessions. Save them through Claude Code's auto-memory system so they show up automatically next time.

This skill is the manual end-of-conversation companion to the auto-memory system already described in your system prompt. Defer to that system as the source of truth — this skill just walks you through running it deliberately at session end. If anything here conflicts with the auto-memory rules in the system prompt, the system prompt wins.

Where memories live

Use the auto-memory directory for the current project — do not hardcode a path. It's the directory referenced as the memory location in the system prompt's auto-memory section (typically ~/.claude/projects/<project-slug>/memory/). The index file is MEMORY.md in that same directory.

If you cannot determine the project's memory directory from the system prompt, ask the user before writing anything.

What to extract (and what to skip)

Worth saving — anything that will help a future session that doesn't have this conversation's context:

  • Corrections — the user told you to stop doing X, or to do Y instead. Save the rule + the why.
  • Validated non-obvious successes — you took an unusual approach, the user confirmed it ("yes, exactly", "perfect, keep doing that"), and a future Claude wouldn't guess this from the code alone. These are easy to miss because they're quiet — watch for them.
  • Project facts with a deadline or owner — merge freezes, who's driving an initiative, the reason behind a rewrite. Convert relative dates to absolute (Thursday2026-05-07).
  • External system pointers — "bugs live in Linear project X", "oncall watches dashboard Y".
  • User profile signals — role, expertise, what framings they prefer.

Do NOT save (these are derivable or ephemeral, even if the user asks):

  • Code patterns, file paths, conventions, architecture — grep/Read the codebase next time.
  • Git history, recent commits, who-changed-what — git log/git blame is authoritative.
  • Debug solutions or fix recipes — the fix is in the commit, the reasoning in the message.
  • Anything already in CLAUDE.md.
  • In-progress task state, current conversation context, today's TODO list.

Read the full file on GitHub · 145 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.

  1. 3d ago First seen · 145 lines · 64 tokens per session scan A 21d3b076239e

Subscribe to this mod's changes

qmemory is a skill published in the GitHub repository 3awny/qship (2 stars, last pushed 2mo ago), licensed MIT. It adds 64 tokens to every session and 1,827 once invoked, about $0.0003 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.

Related

Other skills, from other repositories

maintaining-context

Use at the start of any substantial task and whenever you make a non-obvious decision or learn a reusable fact — keeps .yoke/context/ (PROJECT, DECISIONS, KNOWLEDGE) the durable source of truth so fresh-context work never drifts.

HECer/yoke · 57 tokens

ccc-knowledge

Knowledge compounding — search past lessons, view learning history, browse by category. Use when the user says 'what did we learn', 'past lessons', 'knowledge base',…

KevinZai/commander · 40 tokens

habit

Distill a lesson — a mistake, a user correction, or a praised behavior — into HABITS.md as a positive practice or a hard line. Use after the user corrects you, after an incident, or when the learning-nudge hook fires.

choiyounggi/groundwork · 53 tokens

tutor

Run a spaced-repetition self-quiz over lessons already distilled into HABITS.md — one novel-scenario transfer question per due item, anti-sycophancy grading, and a 1-4 recall rating logged through the tutor scheduler. Use when the user asks to be quizzed or reviewed (English or Korean), or replies to a memory-loop…

choiyounggi/groundwork · 91 tokens

heptabase-cli

Interact with Heptabase using the CLI to create, read, and edit notes, journals, tags, cards, list whiteboards and manage cards on whiteboards, and browse AI Tutor goals, courses, and lessons. Use when the user asks to manage their Heptabase knowledge base, search cards, work with journals, tags, or whiteboards, or…

madeyexz/heptabase-skill · 82 tokens

quick-learning

Owner of the unified methodology knowledge system: format, triad structure, similarity check, Seen counters, and Adapted tracking. Signal-gated: skips clean sessions automatically (zero cost). Automatic trigger: called by feature-execution, do-task, and design-generate (context exhaustion). Use when: "quick learning"…

stepanenkoviktor0110-boop/ai-dev-methodology-codex · 105 tokens