memory-compaction

memory-compaction is a skill for Claude Code, Codex from robotaitai/project-bedrock. It costs 43 tokens per session (863 once invoked), scanned A, original, MIT.

Instructions for compacting and pruning an agent memory tree when its notes become too large, stale or repetitive. It preserves durable facts while reorganising the memory files into a shorter structure.

In plain words
What is it for?
Auditing memory size, merging or moving durable facts, removing superseded notes and maintaining the required memory-file structure.
Why use it?
It keeps project memory easy to scan and removes outdated, speculative or broken references without changing the original evidence files.

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/robotaitai/project-bedrock/memory-compaction
Any agent
npx skills add robotaitai/project-bedrock --skill memory-compaction
Clone the repo
git clone --depth 1 https://github.com/robotaitai/project-bedrock

Made for: Claude Code, Codex.

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 memory-compaction

README.md
[![agentmods](https://agentmods.dev/badge/skills/robotaitai/project-bedrock/memory-compaction.svg)](https://agentmods.dev/skills/robotaitai/project-bedrock/memory-compaction)
Your own site
<a href="https://agentmods.dev/skills/robotaitai/project-bedrock/memory-compaction"><img src="https://agentmods.dev/badge/skills/robotaitai/project-bedrock/memory-compaction.svg" alt="Measured on agentmods" height="20"></a>
Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 863 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 $0.00043 $0.00863
Opus 5 $0.00022 $0.00432
Sonnet 5 $0.00009 $0.00173
Haiku 4.5 $0.00004 $0.00086

Measured 5d ago against content hash 7c71fbb0de09, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

memory-compaction 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.

assets/skills/memory-compaction/SKILL.md · 116 lines

How it starts

The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Memory Compaction

Reorganizes the memory tree to keep it scannable without losing durable facts.

Run this when compact-memory.sh reports warnings, or proactively after dense work periods.


Invariants -- never violate

  • Every durable fact must survive compaction (move or merge, never delete)
  • Memory/MEMORY.md must stay short and scannable
  • Every branch folder must contain a same-name entry note (<topic>/<topic>.md)
  • Evidence files are never touched during compaction

Step 1 -- Audit

Run the report script first:

scripts/compact-memory.sh /path/to/project

Note which files are flagged as:

  • CRITICAL (>200 lines) -- act immediately
  • WARNING (>150 lines) -- split at next opportunity
  • stub or empty -- populate or remove

Step 2 -- Prune stale entries

In each branch note, remove or update entries that are:

  • Superseded: a newer entry contradicts the old one -> remove the old, update the new
  • Dead reference: refers to a path, API, or dependency that no longer exists -> remove
  • Speculation never confirmed: still marked with "might", "probably", "TODO: verify" -> remove or move to Open Questions
  • Recent Changes too old: entries older than ~4 weeks -> move stable facts to Current State, drop transient ones

Do not prune:

  • Gotchas and lessons (even old ones are valuable if the risk still exists)
  • Decision rationale (even for reversed decisions -- mark as Status: reversed)

Step 3 -- Split large branch notes

If a branch note exceeds ~150 lines:

  1. Identify the dominant sub-topic that makes it large
  2. Promote the branch to a folder if it is still a flat note:
    • Create Memory/<topic>/ directory
    • Move Memory/<topic>.md to Memory/<topic>/<topic>.md
  3. Create Memory/<topic>/<subtopic>.md for the extracted content
  4. Replace the moved section in the entry note with a link:
    See [subtopic.md](subtopic.md) for details.
    
  5. Update Memory/MEMORY.md branch link if the path changed

Read the full file on GitHub · 116 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. 5d ago First seen · 116 lines · 43 tokens per session scan A 7c71fbb0de09

Subscribe to this mod's changes

memory-compaction is a skill published in the GitHub repository robotaitai/project-bedrock (45 stars, last pushed 15d ago), licensed MIT. It adds 43 tokens to every session and 863 once invoked, about $0.0002 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 2026-08-30.

Related

Other skills, from other repositories

media-ingest

Ingest video, audio, PDF, book, screenshot, and GitHub repo content into the brain. Multi-format handling with entity extraction and backlink propagation. Covers video-ingest, youtube-ingest, and book-ingest subtypes.

garrytan/gbrain · 52 tokens

mem0-oss-to-platform

Plan and then execute a migration of a project from the mem0 open-source / self-hosted SDK (the local Memory class) to the mem0 Platform / hosted / managed SDK (the MemoryClient class). Use this whenever a developer wants to move, switch, or migrate their mem0 usage off OSS/self-hosted to the hosted API — e.g.…

mem0ai/mem0 · 273 tokens

Cortex

Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…

danielmiessler/LifeOS · 196 tokens

agent-memory

../../../engineering/agent-memory/skills/agent-memory/SKILL.md.

alirezarezvani/claude-skills · 0 tokens

memory

Use when the user asks to remember, recall, forget, update, search, or inspect durable OpenSquilla memory, including profile facts in USER.md and long-term notes in MEMORY.md or memory//.md.

opensquilla/opensquilla · 44 tokens

ha-data-stores

Map of Hope Agent's local data stores and safe read-only query workflow. Use when the user asks where Hope Agent stores data, wants to inspect sessions/messages/memory/logs/background jobs/knowledge indexes/settings, asks the model to query local app data, or debugging requires checking persisted state. Trigger…

shiwenwen/hope-agent · 115 tokens