memory-commit

A memory-saving workflow for storing a specific decision, preference, fact, skill, task, or conversation detail for later retrieval.

In plain words
What is it for?
Use it when someone explicitly asks to remember or save information, or when a deliberate decision or important correction should persist.
Why use it?
It records not only what was said but also why it matters, making future recall more precise.

Skill for Claude CodeCodex

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/n24q02m/mnemo-mcp/memory-commit
Any agent
npx skills add n24q02m/mnemo-mcp --skill memory-commit
Clone the repo
git clone --depth 1 https://github.com/n24q02m/mnemo-mcp

Made for: Claude Code, Codex.

Per session 72 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,065 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 $0.00072 $0.01065
Opus 5 $0.00036 $0.00532
Sonnet 5 $0.00014 $0.00213
Haiku 4.5 $0.00007 $0.00106

Measured 2d ago against content hash 241884a0c81c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

memory-commit 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 2d 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.

Origin

Copies of this mod

2 near-identical copies found in the catalogue:

skills/memory-commit/SKILL.md · 112 lines

How it starts

The opening of the file, as written. The whole thing — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Memory Commit

Manual capture of a single decision, preference, fact, skill, task, or conversation snippet into mnemo. Enforces the "WHY not just WHAT" rule and chooses the correct context_type so retrieval is precise.

When to Use

Trigger on explicit user signals:

  • "remember this", "save this", "save for next time"
  • "ghi nho", "luu lai", "nho lai"
  • "let's commit this to memory", "store this"
  • After a deliberate decision: "we picked X because Y"
  • After a stated preference: "I always want X"
  • After correcting the agent: high-value capture (prevents repeat)

Do NOT trigger on incidental mentions or speculative options not yet chosen.

Steps

  1. Identify the content the user wants to remember:

    • Preceding 1-3 messages (default), OR
    • A specific quoted span the user references, OR
    • The current selection if invoked via slash command
  2. Determine context_type via this decision tree:

    Signal context_type
    "we decided", "we picked X over Y", "going with" decision
    "I prefer", "I always want", "default to" preference
    "X is at version Y", env vars, API shapes, file locations fact
    "to do X you run Y then Z", procedure, how-to skill
    "todo", "remember to", deadline, "by Friday" task
    none of the above (general context) conversation

    When ambiguous, ask the user one short question. Do not silently default to conversation for high-signal content.

  3. Compose the capture text with WHY included:

    • BAD: "Use Polars"
    • GOOD: "Use Polars for dataframes (not pandas) because the codebase processes 50M-row datasets and Polars is 20x faster than pandas in our benchmark."
  4. Capture via the typed action:

    memory(action="capture",
           text="<composed text>",
           context_type="<chosen type>",
           category="<project-name or topic>",
           tags=["<topic>", "<scope>"])
    
  5. Confirm to the user:

    Saved as <context_type>. Memory ID: <id>.
    (Deduplicated against existing similar memory: <existing_id>.)  # if applicable
    

Read the full file on GitHub · 112 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. 2d ago First seen · 112 lines · 72 tokens per session scan A 241884a0c81c

Subscribe to this mod's changes

memory-commit is a skill published in the GitHub repository n24q02m/mnemo-mcp (10 stars, last pushed 2d ago), licensed Apache-2.0. It adds 72 tokens to every session and 1,065 once invoked, about $0.0004 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.